AI-Decision Making: State Of Play And What’s Next

AI alone was not up to the job so Finland’s largest airline instead implemented a hybrid system that uses AI to make predictions about air traffic and allows the humans-in-the-loop to make better decisions, explains Tero Ojanpera, CEO of Silo.ai, a Finnish AI lab that specializes in bringing cutting-edge AI talent to corporations around the world.

Getting the FinnAir project to that point was not a question of plug and play. It required a complex multi-step modeling process to help the organization become more AI literate.

Finnair’s experience neatly illustrates the current state of play. AI is not fully ready to make the kind of decision-making corporates expect it to make and even if it were corporate teams and networks are not fully ready to implement and reap the full benefits of AI.

The state of AI-decision making was the focus of an October 13 roundtable discussion moderated by The Innovator in partnership with DataSeries, a global network of data leaders led by venture capital firm OpenOcean. The discussion centered on what is holding back business from using AI, how corporates should approach AI projects in order to better leverage the technology and the methods being tested to improve AI’s decision-making powers. Some of these methods – such as the merger of rules-based and machine learning (ML) techniques, knowledge graphs and multi-modal neural sequencing – promise not just to help automate existing functions but to aid companies to strengthen and even re-imagine their businesses.

Stumbling Blocks

There are many variables at play for a successful AI product to go from infancy to launch.

“Corporates need to make sure their whole infrastructure is ready before trying to build something more intelligent on top,” says roundtable participant Ekaterina Almasque, a general partner at OpenOcean. “Unfortunately, in many enterprises there is still the question of how to deal with the data.”

She cited the example of the automotive industry, which is searching for new sources of revenue, such as using AI to leverage data collected from connected cars. The automakers don’t even have data centers that can collect and process the data in a way that it can be used, she noted, creating an opportunity for startups to help them develop ways to close that gap.

Other corporates have been collecting lots of data for years. But when they start on AI projects they sometimes find that a few columns of crucial information are missing. “It doesn’t matter how big the data is or how long it has been collected, it is not necessarily the perfect data,” says round table participant Reza Khorshidi, one of the founders of, and currently a research leader at, the Deep Medicine Program at the University of Oxford’s Martin School, and also the Chief Scientist for global insurance company AIG.

Corporates don’t only need to ensure that they have the right data – and enough of it – regardless of whether it comes from different parts of the organization or from a variety of outside sources. The data also needs to be structured properly, which is no easy feat.

“If there is one thing that we could do to save hundreds of billions of dollars every year it is to start with standardization of data schema,” says round table participant Vishal Chatrath, CEO of Secondmind, a U.K. startup that helps corporates identify and address how to to build a decision-making framework that combines the best of AI with human domain knowledge.

A lack of data compatibility makes it impossible to businesses to fully leverage AI insights and for supply chains to operate efficiently. Chatrath used the example of an online shop in the UK that sells branded sport t-shirts. But there is no way for the shop to alert Nike or Adidas that the green t-shirt is out of stock and proactively order green dye, buttons and thread, because there is no universally recognized way to call a green t-shirt a green t-shirt. This type of problem was avoided when the mobile Internet was created because a lot of effort was put into standardization, resulting in a Global System for Mobile Communications (GSM). “We need the equivalent of a GSM for AI,” says Chatrath. “Someone has to take this bull by the horns and say ‘dammit you have to standardize data schema’.”

To get the best out of AI, corporations should count on spending about 80% of their time putting into place the digital foundations, says roundtable participant Simon Greenman, co-founder and a partner at Best Practice AI, a U.K.-based management consultancy that specializes in helping companies create competitive advantage with AI. “AI is actually the easy bit,” says Greenman. “The hard bit is making sure the organization really has the platforms and the technology in place to be able to do AI. These things are slowing down the adoption curve.”

Clearly the best way to create systems that are more intelligent is to find the right data or “generate the data you need in a synthetic way to build your models and train them,” says roundtable participant Jose Luiz Florez, an AI expert and founder of a number of startups, including Dive.ai. “But if your models are not good enough then you need to put humans-in-the-loop.”

Getting Ready To Launch

That’s exactly what Finnair ended up doing. The expectation was that AI could actually decide what actions to take if congestion was increasing but AI is currently not able to handle such multi-dimensional optimization problems, says roundtable participant Ojanpera, who previously worked as Nokia’s Chief Technology Officer. Silo, his current company, helped Finnair develop a model to more accurately predict 36 hours in advance how many planes would be delayed based on various factors, such as bad weather. That was just a start because it is only one piece in a complicated problem, he says. When companies deploy such solutions they need to factor in a way to ensure that the human decision makers understand how the model works so that they actually believe what the model says. Then, and only then, can they start to look at the next problem: the possibility of automating some of the decisions that follow when AI starts to understand the context and the situation better. “It is important to break down the problem into pieces and start by selecting the one that will produce the best output in the short term,” Ojanpera says. “That’s how organizations become more AI literate. They start to understand what AI can and can’t do, and I think that’s a good starting point.”

In addition to ensuring the necessary data is ready and the technology underpinning is up to the job, corporates need to invest in AI talent, says AIG’s Khorshidi. “Don’t think by some magic trick that your company is going to go from pre-AI to AI-first without it,” he says. Once the team is in place a system needs to be put in place to properly test the AI and some sort of domain expertise is needed. There are a number of options for adding this expertise, including relying on employees’ insights.

“Usually we talk about data in the traditional sense but it is important to remember that many industries and businesses have been run by experts,” says Best Practice AI’s Greenman. “These human experts are a form of data – they have cognitive data that no company has managed to capture yet with tools. So, if companies are starting to collect more data they definitely should start looking more at their employees and use that cognitive data to help machine learning models to get better and better over time.”

Keeping Humans-In-The-Loop

A number of the roundtable participants argue that – for the time being – hybrid systems are the best if not the only real option to obtain better AI decision-making. “There is a need for humans in-the-loop and I don’t see that going away anytime soon,” says roundtable participant Chatrath, Secondmind’s CEO.

The Cambridge-based startup has developed what it calls the Secondmind Decision Engine, a machine learning-powered software-as-a-service platform designed to aid decision-making across industries, including, it says, “those in which visibility is low, data is sparse, and uncertainty is high.”

Currently in limited release, Secondmind’s Decision Engine is already being used by Kuehne+Nagel, a sea logistics provider which coordinates the movement of nearly 13,000 shipping containers per day and Brambles, an Australian company that specializes in the pooling of unit-load equipment, pallets, crates and containers. The companies are using the startup’s technology to make demand forecasting, planning and asset allocation decisions within their global supply chain operations.

Secondmind says its technology can offer – on average – a 35% improvement in efficiency by using a combination of Gaussian Process-based probabilistic modelling and decision-making machine learning libraries. The technology suite is adept at quantifying uncertainty, identifying operational trade-offs and explaining outcomes using sparse and low volume data, capabilities that meet business decision-making demands where other machine learning techniques like Deep Learning struggle, says Chatrath. Still, he says, humans with industry knowledge are crucial to success.

That’s a viewpoint that was shared by round table participant Jinsook Han, Accenture’s Managing Director and Global Lead of Growth and Strategy, Applied Intelligence. “Having humans in the loop is extremely important regardless of whether you are thinking about AI as a means of augmenting, accelerating, assisting and I would add a fourth “a” for avoid [ looking at the technology as a means of reducing risk], says Han. When chatbots were first being introduced clients came to Accenture and asked whether they could get rid of all 1000 employees in their call centers, she says. “We told them that is not the way you want to go. Let’s focus on what is important for the customer experience. What if the call center rep had enough information from the AI when they picked up the phone to be able to resolve the problem on the spot and give the customer a better experience? There are times when AI can handle an issue and other times when it is preferable to have a human-in-the-loop. My mantra is that we should let humans do what humans do best and let machines do what machines do best. Clients are beginning to understand that this is a journey.”

New Approaches

While corporates try and get their own houses in order data scientists are working on a number of ways to improve AI-decision making.Rules-based approaches, which use structured data, have been used to make intelligent business decisions since the 1980s. When you are dealing with multi-modal messy data and more complex problems it is generally agreed that machine learning is the better approach. But it is far from perfect. If new regulations come into play or the rules of the past no longer apply, ML, which has been trained on historical data, has no clue what to do – or even to recognize that the context has changed. “There is a lot of work being done now on how to correct these problems without losing the value derived from ML,” says Harley Davis, head of IBM France’s R &D Lab.

Merging Rules-Based Systems And Machine Learning

One way to try and solve for this is to merge rules-based systems with ML. IBM launched a new product in 2020 called Automation Decision Services that does just that. Combining the two approaches leads to better decision making, says Davis.

IBM’s Automation Decision Services product and its predecessor, IBM Operational Decision Manager, are now being employed in a variety of sectors, including financial services and aviation. For example, all of PayPal’s transactions now use a combination of ML fraud detection and explicit business rules to identify some very specific security concerns, says Davis. Mastercard is using it to do something similar – using the combination of ML analytics along with business rules developed with 800 member banks to detect fraud. In the U.S. Fannie Mae and Freddie Mac (federally-backed home mortgage companies created by the U.S. Congress) are using IBM’s technology to process over two-thirds of US mortgage applications, using ML on top of symbolic representation of rules, to make better decisions, he says. And airlines, such as Delta and United, are using it to figure out the best way to create upgrade offers. They are also using logic-based programming known as mathematical optimization to deal with complex logistical issues such as rescheduling. This approach – which has been around for decades in operations research – is now also being combined with AI-based predictions, in an IBM product called IBM Decision Optimization, says Davis, though he concedes that there is still often a need for humans-in-the-loop.

Knowledge Graphs

Knowledge graphs provide another way to start emulating the implicit functions of the human mind and combine it with the computing power of machines to represent meaning by putting data into context, similar to the way humans connect pieces of information to reach a conclusion. They are being used in Alexa and Siri voice assistant devices and in Google searches and they are starting to be applied in different industries, such as pharmaceuticals, chemicals R&D and oil and gas, using an approach IBM calls cognitive discovery.

IBM Research says it has developed a scalable pipeline of technologies that can be leveraged to extract information from highly unstructured sources such as documents or scanned images. This is typically the form in which companies have stored their knowledge and experience. For example, when scientists publish their papers and patents it is can be hard to process digitally. Cognitive discovery aims to automate the extraction of knowledge from these ‘dormant’ sources, combine it with other structured and semi-structured information and make it readily available via a user-friendly interface on top of a knowledge graph. This rich body of knowledge preserves the corporate wisdom and experience and makes it readily available for research and development in specific industries, says Stefan Mueck, IBM Germany’s Chief Technology Officer responsible for digital transformation.

This approach constitutes a new, accelerated and better way of doing R & D, he says. Leveraging the data pipeline for both internal and external sources (e.g. patents) researchers can start building a hypothesis based on a much bigger and broader input than any human could possibly read or have present in his mind. To help him even further, AI could be used to infer knowledge from what is in the graph.

IBM cites the following example, based on real-world experience with a number of chemicals companies as well as in its own material science research: Say a chemicals company wants to create a new and innovative material or substance with certain properties. Given all the internal knowledge plus all the external information that is free or licensed for use, the machine would extract what it finds about ingredients, formulations and processing as well as the related product properties of the end product. IBM Research has developed deep learning models that support the researcher with predictions: Given a formulation and a process, what would be the properties of the new substance? Or, given the properties, what would be the formulation? Similarly, if it is about base chemical reactions there are models to predict synthesis or retro-synthesis. The latter is made open source by IBM. It helps companies build extended capabilities on top of it. The decision intelligence can be driven even one step further by integrating predictions with lab automation, which is one of the latest innovations. This cognitive discovery capability has also been made available to the public for accelerating research around treatments or finding vaccines for Covid-19. Other industries are also using knowledge graphs to improve AI decision-making. The use of IBM’s cognitive discovery is under development with “an innovative player in the oil & gas sector,” says Mueck. It will be revealed at an industry conference called EAGE Digital in November, he says.

IBM believes knowledge graphs can offer business another big advantage. In some -but not all -cases, adding knowledge graphs to a combination of ML and rules-based systems can help companies explain why an AI made a particular decision, helping resolve serious social, legal and ethical concerns. “Machine learning represents a big black box. We don’t know why it gets the results it does which introduces multiple social, ethical and legal problems,” says Davis. “By looking at the knowledge graph and the rules you can give an explanation for a decision like why a loan was turned down: it was because your credit history was bad and your revenue was insufficient and so forth,” he says.

That said, there are still a number of tough problems to solve before more automated decision-making systems can be more widely and safely used by business, says Davis. He cites the well-known case of Amazon having to redesign an AI system that developed for recruiting purposes because it found that it was biased against women candidates. Amazon removed names and gender from the applications but other indicators such as hobbies or schools still led to the AI to give better ratings to male candidates, because previous human decisions favored males and those biases were correlated with other data in the resumes. There are tools that can find those biases, for instance IBM OpenScale, but you have to know what to look for and run analysis on the ML training data and it is not an easy problem to solve.

Decision Intelligence

New types of approaches – including the social sciences – may need to be introduced into ML models, says Davis. That’s where decision intelligence comes in. The term – which made it into Gartner’s 2020 hype cycle – refers to an emerging engineering discipline that augments data science with theory from social science, decision theory and managerial science to try and provide a framework for best practices in organizational decision-making and process for applying machine learning at scale. Gartner has developed a Decision Intelligence Model to help business executives identify and accommodate uncertainty factors and evaluate the contributing decision-modeling techniques.

Transformer-based Sequence Models

While all of these approaches may help businesses move closer to AI-led decision-making roundtable participant Khorshidi believes Transformer-based neural sequence models, which have shown tremendous advances in natural language processing, have the best opportunity for success. If these models can be tweaked to accommodate the multimodal nature of data, he believes “it will have the ability to go beyond language, beyond health, beyond finance and pretty much cover every real-world data generating process,” he says, helping to transform business as we know it.

Khorshidi and his team at Oxford’s Deep Medicine Program have had success testing the application of Transformer-based models to sequence the multi-modal biomedical data found in electronic health records.

Electronic health records are sequences of mixed-type data such as diagnoses, medications, measurements, interventions and more that happen in irregular intervals and are routinely collected by health systems.

If Khorshidi and his team’s initial positive results, which were published in Nature magazine last April, can be replicated at scale and across a range of data scenarios “this could mean a breakthrough in medicine and be the difference between pre-AI/AI-inside medicine and AI-first medicine,” he says. The breakthrough is tied to the ability to build complete electronic health records that include what health systems have been routinely collecting, as well as social, economic, environmental and lifestyle data that have been shown to be important (and predictive) for health outcomes. A sequence model’s ability to learn the key patterns and relationships/dependencies underlying such complex sequences will enable health systems’ ability to anticipate things before they happen, and intervene when needed. “This can enable better, cheaper, faster processes, and ultimately pave the way towards redesigning and re-imagining the system,” Khorshidi says. The same approach could also be applied to other sectors, such as finance and retail, he says.

“Dealing with sequential data that is mixed-type multimodal and happens in irregular intervals gives machines the ability to deal with any sort of data,” says Khorshidi. “In the real world it could be a customer’s data on Amazon, it could be a patient’s data in NHS or it could be a company’s data for asset management.”

If not just medicine but other industry sectors want to move to a world of high dimensional data in which the data is inputted in a messy way, there are not many solutions out there,” says Khorshidi. “You need to settle on some sort of feature engineering or settle for models that can deal with data as they arrive sequentially. And that’s why I think transformer based architectures have got higher odds of success.”

Regardless of which new method businesses use to improve AI decision making they should change the way they think about AI, he says.

“The true north for AI is transformation opportunities and reimagination opportunities,” Khorshidi noted at the end of the roundtable. “Automation is the lowest hanging fruit. We should use AI to reimagine the power of the existing base of employees and strengthen businesses by doing things differently.”

Rebooting Copyright For The Age Of AI

 

“Is This What We Want?” is the name of a silent album released by UK musicians, including Annie Lennox and Kate Bush, to protest UK government plans to allow AI companies to use copyright-protected work without permission. It is just one of the many ways that well-known figures from the publishing, music, film, TV, design and performing arts sectors are displaying their displeasure over proposed changes to the country’s copyright law.

Data is the lifeblood of artificial intelligence, and large language models – such as ChatGPT – are training on vast amounts of publicly available data sets. They are reeling in content on the Internet produced by musicians, journalists, artist and authors, ballooning their own valuations while threatening the livelihoods of content creators.

Copyright lawsuits filed against GenAI companies abound, alleging that the way they operate amounts to theft. Examples include the New York Times v OpenAI lawsuit in the U.S., and in the Getty Images v Stability AI case in the UK. The allegations in AI and copyright cases generally split into two parts: first, that the outputs of AI models constitute an illegal copy; second, that using copyrighted works in training data for AI (inputs) is a breach of the image owner’s copyright.

Requiring developers to license all the material they use to train models would be very difficult due to the distributed nature of data and ownership, argues the Tony Blair Institute for Global Change (TBI). To provide legal clarity and accelerate AI development, some countries have already taken a lenient view on use of publicly available data for AI training, so if other countries take a restrictive stance, it will drive development elsewhere. And even if a workable solution for payment were found it would likely stifle competition since only large, well-funded AI companies could afford to pay.

“As technologies and societies evolve so must regulations,” Jakob Mökander, TBI’s Director of Science & Technology Policy said in an interview with The Innovator. “We need to find an approach that makes sense in the digital age. It is a question that all governments will face.”

On April 2 TBI published a report on rebooting copyright that says the current situation is unsustainable. It argues that the status quo harms all stakeholders, including creators, who are not properly remunerated for their labor; rights holders, who struggle to exercise control over how their works are used; AI developers, who face hurdles when it comes to training AI models, and society at large, which risks missing out on benefiting from AI diffusion and adoption. “Bold policy solutions are needed to provide all parties with legal clarity and unlock investments that spur innovation, job creation and economic growth,” says the report.

The TBI report supports the position favored by the UK government: a text and data mining (TDM) exception for AI model training with the possibility for creators and rights holders to opt out. This would make it legal to train AI models on publicly available data for all purposes, while giving rights holders more control over how they communicate their preferences with respect to AI training, argues the report.

The report notes that it is important to separate the debates around AI outputs and AI training. AI outputs should not be allowed to reproduce original works without proper license and remuneration, says the report, but prohibiting AI models from training on publicly available data would be misguided and impractical. “The free flow of information has been a key principle of the open Web since its inception,” says the report. “To argue that commercial AI models cannot learn from open content on the Web would be close to arguing that knowledge workers cannot profit from insights they get when reading the same content.”

There are better ways of supporting the creative industries, says Mökander, “There needs to be increased funding for creators in the digital age, we want to have flourishing industries, but copyright law may not be the best way to do that,” he says

The report suggests some alternative approaches to helping the creative industries, including the creation in the UK of a Centre for AI and Creative Industries which would serve three functions: bringing together experts and representatives; acting as an engine to create new technologies and infrastructures to support growth in machine learning in the UK creative industries; and providing much-needed training and expertise across academia and industry. If more funding for the arts are needed, and if governments need to raise more funds for this, one option to consider is taxing data connections on fixed lines and mobile devices, adding pennies per month to the Internet Service Provider (ISP) bills of households and businesses who benefit from using AI tools.

“Rather than fighting to uphold 20th-century regulations, rights holders and policymakers should focus on building a future where creativity is valued and respected alongside AI innovation,” says the report: “Copyright law provides insufficient clarity for creators, rights holders, developers and consumer groups, impeding innovation while failing to address creator concerns about consent and compensation. The question is not whether generative AI will transform creative industries (it already is) but how to make this transition equitable and beneficial for all stakeholders.”

The Trouble With Copyright

The truth is that no one is happy with the status quo, says Mökander.

Just ask American author Cory Doctorow. “For 40 years, the scope and duration of copyright have monotonically increased, the evidentiary burden for copyright claims has declined, and the statutory damages for copyright infringement have expanded,” he wrote in an online article.  “Publishing and other creative industries’ generate more money than ever – and yet, despite all this copyright and all the money that sloshes around as a result of it, the share of the income from creative work that goes to creators has only declined. The decline continues. There is no bottom in sight.”

Doctorow uses the following analogy to drive home his point: “If the bullies at the school gate steal your kid’s lunch money every day, it doesn’t matter how much lunch money you give your kid, he’s not gonna get lunch.  But how much lunch money you give your kid does matter – to the bullies. (Creators) are the hungry schoolkids. The cartels that control access to our audiences are the bullies. The lunch money is copyright.”

Strengthening copyright law would do little to benefit creators and requiring developers to license the materials needed to train AI would  threaten the development of more innovative and inclusive AI models, as well as important uses of AI as a tool for expression and scientific research, argues the Electronic Frontier Foundation (EFF), which has published a series of articles on problems with copyright in the age of AI.

Requiring researchers to license fair uses of AI training data could make socially valuable research based on machine learning and even text and data mining  prohibitively complicated and expensive, if not impossible, argues the EFF. It notes that researchers have relied on fair use to conduct TDM research for a decade, leading to important advancements in science and other fields.

For giant tech companies that can afford to pay, pricey licensing deals offer a way to lock in their dominant positions in the generative AI market by creating prohibitive barriers to entry, says the EFF. To develop a foundation model that can be used to build generative AI systems like ChatGPT and Stable Diffusion, developers need to train the model on billions or even trillions of works, often copied from the open Internet without permission from copyright holders. There’s no feasible way to identify all of the rights holders—let alone execute deals with each of them. Even if these deals were possible, licensing that much content at the prices developers are currently paying would be prohibitively expensive for most would-be competitors.

As the U.S. Federal Trade Commission recently explained, if a handful of companies control AI training data“they may be able to leverage their control to dampen or distort competition in generative AI markets” and “wield outsized influence over a significant swath of economic activity.”

The Way Forward

The UK’s proposal for a TDM exception with opt-out – which essentially allows the scraping of publicly available information- would bring UK regulation broadly in line with the European Union’s.

But other jurisdictions, such as Singapore and Japan, have more liberal copyright laws pertaining to AI training and China is speeding ahead. The current administration has indicated that the U.S. will not pursue strict AI regulations but there is ongoing litigation in the U.S. around AI training. What constitutes fair use of copyrighted materials in the U.S. will be decided on a case-by-case basis.

The legal landscape surrounding IP data scraping is not only complex it is rapidly evolving, says a February OECD report on data scraping. What’s more different actors in the data scraping ecosystem raise various types of legal issues. Some also use data scraping to support research and other endeavors, suggesting the need for policy tools tailored to different use cases, says the OECD report. The data scraping ecosystem encompasses research institutions and academia, AI data aggregators, as well as technology companies and platform operators. Research institutions and academia frequently employ data scraping to gather data for academic and scientific purposes. AI data aggregators make scraped data available to third parties, often without clear licensing terms or clear disclosure of data provenance, raising IP and other legal concerns. Technology companies and platform operators are sources of scraped data and regular data scrapers themselves.

The OECD is promoting a global data scraping code of conduct, standard contract terms, standard technical tools and initiatives for building awareness that would chart a responsible path for data scraping in an internationally coordinated manner. “This would be particularly effective if it is developed with input from a broad and diverse set of stakeholders, including rights holders, researchers, AI developers, civil society, and policymakers,” says the OECD report.

TBI’s Mökander says he would welcome  globally recognized codes of conduct. ”AI training data are only useful if there are clear international standards,” he says. “ If not, we risk a race to the bottom, pushing AI development to other jurisdictions with more lenient regulations. In fact, harmonized international standards should be top priority for policymakers seeking to build a flourishing ecosystem for the arts and AI.”

Tech tools will help ensure compliance, he says. For example, AI company Spawning has developed a Do Not Train registry that allows artists to tag work around the Internet as copies of their original. Developers can then use “data-diligence” software from Spawning to check whether URLs have been opted out.Another tool cited in the TBI report is ProRata.ai, a new company that uses tech to enable generative artificial intelligence (GenAI) platforms to attribute and compensate content owners.

ProRata CEO Bill Gross invented tech that can reverse-engineer where an answer came from and what percentage comes from a particular source so that owners can be paid for the use of their material on a per-use basis, says Gross, who has patented the technology. ProRata pledges to share half the revenue from subscriptions and advertising with its licensing partners, help them track how their content is being used by AIs, and aggressively drive traffic to their websites.

When a user poses a query ProRata’s algorithm compiles an answer from the best information available. At the top of the page there is an attribution bar which specifies where the answer came from. It might say, for example, 30% of this answer came from The Atlantic, 50% from Fortune and 20% from The Guardian. The publications are immediately compensated according to their contribution to the answer and a side panel displays the original articles and enables users to click on the original source to learn more. Think of it as “attribution-as-a-service.,” Gross said in an interview earlier this year with The Innovator. “Just as Nielsen measures how TV shows are watched to determine what advertisers should pay, we are moderating the output of the queries to determine how much GenAI providers should pay content providers.”

In the future it should be technically simple to build AI agents that can track creators’ portfolios, maintain registries and initiate robot-to-robot interactions with other websites, asking them to remove content, says the TBI report. These agents are expected to simplify content attribution for AI companies, enabling them to effectively track online content origins and eliminating plausible deniability for developers who claim ignorance about opted-out work appearing in their systems.

If the right policies are put in place the AI revolution “can be the standout engine for artistic and cultural renewal of our era,” says the TBI report. It could also help countries like the UK lead in the AI sector.

Time To Act

But time is not on anyone’s side, says the TBI report.  Large, effective foundation models already exist and are publicly accessible. They will continue to grow in capability and will be used by an increasing number of people. They will be developed around the world, in jurisdictions with very relaxed copyright laws, and used as tools to the extent that they will inevitably make some jobs redundant. At the same time, countries with restrictive laws will push developers to move to countries with less stringent measures, says the TBI report. “The longer governments take to tackle the issue of AI and copyright, the more they will inhibit innovation and entrench large AI developers in the global competition for AI leadership.”

Visit: https://innovatorawards.org/

Ten hard-won lessons from a decade of mobility innovation

After a decade of supporting mobility innovation, the Advanced Propulsion Centre UK (APC) and Zenzic have worked with hundreds of start-ups and SMEs, helping them take their innovation from concept to commercialisation. We’ve seen the breakthroughs and the breakdowns – the big deals and the cautionary tales. Here, Joshua Denne, Head of Product at APC, and Mark Cracknell, Programme Director for Zenzic, reflect on the top ten lessons that separate the companies that accelerate from those that stall.

The mobility sector is brutal. We are heavy on hardware, and unlike software, you can’t pivot on a whim. The product lifecycle is longer, the capital burn is higher, the asset requirement is intense, and proving technical viability is only half the battle. Commercial traction is what defines winners.

We launched the next evolution of our mobility start-up accelerator ‘Mobilise’ in January 2025. These ten hard-won lessons from the past decade have shaped how we hope to support the next generation of mobility start-ups.

1. The right support at the right stage

The UK’s mobility ecosystem is rich in its diversity: we have global corporate OEMs and suppliers, established supply businesses, small-and-steady SMEs, high-ambition scaleups, and bleeding-edge start-ups. Taking a single approach to supporting each of these segments does not deliver the best outcomes. A start-up taking its core technology to Minimum Viable Product (MVP) needs fundamentally different help than one scaling manufacturing capacity. A global OEM does not need the same intervention as an established UK-based supplier. Our focus in this article is on start-ups and the programmes we have developed to accelerate them.

For start-ups at the seed stage, funding for technology validation is immediately critical and an entrepreneur’s first priority. However, due to long product introduction periods, the high cost of development, and the challenging commercial environment, commercialisation expertise, IP strategy, and Investment readiness also need to be an early priority – support with the target market, and help preparing for relevant early-adopter market requirements are absolutely critical.

Mismatched support can leave a lot of value on the table. We’ve seen companies attempt to deliver large application readiness programmes before they’ve validated their technology, only to burn through cash with limited traction. Likewise, we have seen many start-ups waste time and resources shooting for an unrealistic market segment. The right support at the right stage accelerates, but the wrong support at the wrong time can be a distraction.

Mobilise is a structured early-stage accelerator programme that supports ambitious start-ups, university spinouts, or pivoting SMEs that are developing innovative mobility-related early-stage, zero-emission or Connected and Automated Mobility (CAM) technologies, products, services, or solutions to accelerate the transition to a safer, smarter, more sustainable future.

2. Early adopters beat ‘build it and they will come’

Proof of traction trumps proof of concept. It’s easy to fall into the trap of thinking that superior tech will automatically attract buyers. It won’t. Companies that spend years perfecting their product, without bringing early adopters on board, often fail. Likewise, focusing on large multinational customers as innovators or early adopters is a fatal error. We can think of just a handful of companies that have converted commercial deals with global multinationals as their first or early adopters.

The winners engage potential customers early. They focus on initial customer segments that can allow market entry at pace, ideally at a premium, even if the total market size is smaller. In an ideal world, they go beyond letters of intent (LOIs) to secure paid pilots and joint development agreements before scaling. These commitments provide validation and create customer pull, making the eventual commercial launch far less risky.

We’ve seen companies with inferior technology win market share because they had early adopter buy-in. Meanwhile, technically superior start-ups struggle because they waited for the ‘perfect product.’

3. Redefining MVP in hardware: Segment, model, product

The classic concept of a Minimum Viable Product (MVP) doesn’t always translate perfectly to hardware or deep tech. You can’t really ship a half-baked prototype in a highly regulated market. Instead, we think in terms of:

• Minimum Viable Segment: Proving the value in a specific niche (e.g., low-volume EVs before targeting mainstream OEMs), which is big enough to make sense for initial product development, and small enough and innovative enough to be your first customer.
• Minimum Viable Business Model: Demonstrating through a focused go-to-market strategy, developing the minimum viable asset set to service your identified first customer segment.
• Minimum Viable Product: A product with just enough functionality to attract customer traction (including regulatory requirements) of this first segment.

For mobility start-ups, an MVP ≠ prototype. It’s about proving an initial business model, of which your product is just one part.

 Visit us: innovatorawards

The Next Industrial Revolution

AI is helping automate factories and connect the $25 trillion global product economy.

The Industrial Internet of Things will connect an estimated 50 billion assets such as machines, turbines, vehicles and rolling stock from the transportation, energy, healthcare, automotive, manufacturing, mining, oil and water industries. It is part of the next industrial revolution, in which the Internet of Things (IoT), combined with artificial intelligence, will disrupt end-to-end value chains.

Corporations are scrambling to prepare for this restructuring as the scope of change will compel many manufacturers to adopt new plant designs, reshape their manufacturing footprints and devise new supply-chain models. The new approach to production will use machines linked through the Internet to assemble parts and adapt to new processes with minimal guidance from human operators. Siemens’ Electronic Works in Amberg, Germany, is an example of what intelligent manufacturing will look like. There, employees manage and control the production of programmable logic circuits through a virtual factory that replicates the factory floor.



Via bar codes, products communicate with the machines that make them, and the machines communicate amongst themselves to replenish parts and identify problems. Nearly 75 % of the production process is fully automated, and 99.99988% of the logic circuits are defectfree, notes a June McKinsey report.

The report predicts that going forward, the AI’s intelligent insights will help manufacturers across different industries shorten development cycles, improve engineering efficiency, prevent faults, increase safety by automating risky activities, reduce inventory costs with better supply-and-demand planning, and increase revenue with better sales-lead identification and price optimization. The application of AI is also expected to allow end-to-end real-time visibility into the supply chain. “Machine learning will help huge operations get a better understanding of how external upstream and downstream parties interact with their own facilities and functions,” says Mikhail Nader, CEO of Ethereum, an American startup that uses natural-language processing and machine learning to extract and categorize information from more than 10 million supply-chain signals every day. The intersection with AI enables supply-chain participants to be notified
immediately of relevant opportunities or potential disruptions.

“The way we handle our operations will look completely different,” he says. “We will no longer be making decisions based on blind guesswork. By having a better, broader, and more accurate understanding of how and when things are happening, AI-enabled systems can give users the best options, instead of inundating them with irrelevant and non-targeted data.”
A Single Global Supply Chain

That is not all. Currently, companies operate as if they own their own supply chains. “We believe this is a fallacy and that there is only one global supply chain, and this is inherent in the way we are building our software,” says Nader. Ethereum is developing a product that aims to map every supplier, warehouse, factory, and port in the world to provide a full picture of the $25 trillion global product economy. AI is a layer that sits on top of this view of the global supply chain to make sense of the signals coming in, giving users targeted insights.

A well-functioning supply chain is the backbone of virtually every industry, and accurate projections for just the right amount of inventory are critical to achieving a competitive advantage, notes the McKinsey report. Factors such as product introductions, distribution network expansion, weather forecasts, extreme seasonality, and changes in customer perception or media coverage can severely affect the performance of the supply chain, says the report. What’s more, traditional systems for forecasting and replenishment can’t take advantage of the amount of data associated with IoT devices and the sheer number of influencing factors. So supplychain leaders are starting to realize the advantages of applying AI to increase forecasting accuracy, the report says.

Using AI to predict demand is also expected to allow businesses to optimize their sourcing more broadly, including fully automating purchases and order processing. The report cites the example of the German online retailer Otto, which uses an AI application that is 90% accurate in forecasting what the company will sell over the next 30 days. The system has proven so reliable that Otto now builds inventory in anticipation of the orders AI has forecast, enabling the retailer to rush deliveries to customers and minimize product returns. Otto is confident enough in the technology to let it order 200,000 items a month from vendors with no human intervention. AI is not just helpful for forecasting demand for current products. It can replace humans through automation and enable people and robots to work safely side-by-side in factories. For example, in a warehouse of the British online supermarket Ocado, robots steer productfilled bins over a maze of conveyer belts and deliver them to human packers just in time to fill shopping bags. Other robots bring the bags to delivery vans, whose drivers are guided to customers’ homes by an AI application that picks the best route based on traffic conditions and weather.

Advances in computer vision are behind many developments in these collaborative and context-aware robots, according to the McKinsey report.
Enhanced vision is enabled by more powerful computers, new algorithmic models, and large training data sets. Within the field of computer vision, object recognition and semantic segmentation — the ability to categorize a particular object type, such as distinguishing a tool from a component
— have recently advanced significantly in their performance. These changes allow robots to behave appropriately for the context in which they operate. Context-aware robots recognize the materials and objects they interact with and are capable of safely interacting with the real world and with humans.

New AI-enhanced, camera-equipped logistics robots can be trained to recognize empty shelf space. Deep learning can also be used to correctly identify an object and its position, enabling robots to handle objects without requiring things to be in fixed, predefined positions. AI-enhanced logistics robots are also able to integrate disturbances in their movement routines via an unsupervised learning engine for dynamics. Despite such notable progress and the promising potential of AI’s application to factories and the supply chain, industry players have not yet fully embraced the interconnectivity of machines and sensors and the use of data and analytics. Concerns include the potential for the Industrial Internet of Things to act as a conduit for cyber-security attacks.

Visit: https://innovatorawards.org/

How A New Entity Aims To Help Europe Gain AI Sovereignty

Aleia, a new umbrella organization that will federate the best European AI startups into a single centralized unit and present their offerings as a one-stop-shop for corporate clients, launched on February 3 with financial support from an arm of the French government and several angel investors. The aim is to ensure Europe’s technology sovereignty, increase its competitiveness, and make it easier for large corporates to move from proof-of-concept trials to scaling artificial intelligence across their organizations.


The launch event at Aleia’s headquarters in Paris’ 8th arrondissement, included Renaud Vedel, the head of France’s national AI strategy, two former members of the EU’s High Level Expert Group on Artificial Intelligence, the CEOs of several startups and the head of France’s new Cyber Campus, a cybersecurity initiative from French President Emmanuel Macron, which will open its doors this month. The event was moderated by Jennifer L. Schenker, The Innovator’s editor-in-chief.

The €8 million in financing will help Aleia, which is now officially part of France’s national AI strategy, accelerate its development ahead of its commercial launch in June of this year.

The French government’s backing of Aleia is part of a trend. At a time when there is growing mistrust in the handling of data by U.S. and Chinese actors, there is a growing desire in Europe to reduce dependence on foreign technology. The lack of European Cloud services has received much attention and the European Commission’s February 2020 proposals on data and AI stressed the importance of industrial data as a resource that -if leveraged – could give Europe an edge. The notion of sovereignty also intersects with the long-time European concerns about privacy and personal data. Some believe there is an opportunity for Europe to differentiate itself with a brand of AI that is more human-centric, transparent, and trustworthy, incorporating European values, such as data privacy, that could serve as an alternative to American and Chinese offers.

If Europe plays its cards right and develops and diffuses AI it could add some $2.7 trillion, or 19% to its economic output by 2030, according to a report by the World Economic Forum and McKinsey. At stake is not just the competitiveness of nations but the viability of Europe’s largest companies.

Today the uptake of AI by both business and government in Europe is still limited. Many attempts never get beyond proof-of-concept trials. Aleia aims to change that by ensuring that governments and companies have the right data and the right tools to fully leverage the power of AI, says Antoine Couret, Aleia’s CEO. Its ambition is to offer a platform for development that unifies, secures, and industrializes the whole data and AI production chain, from ideation to scaling it across the enterprise. “Think of it as a fast track to AI, both in terms of business impact and tech simplicity” he says.

Europe does not have its own version of an Amazon Web Service, Google or Alibaba. What it does have is best-of-breed startups in different areas of AI, says AI expert Francoise Soulié, a speaker at the launch event. Europe’s AI companies are still relatively small, and their offers are niche, so they struggle to get contracts with big companies or public administrations, she says. By grouping them together on one platform it will be easier for them to compete against offers from foreign tech giants, says Soulié, who is a scientific advisor to France AI, a former member of the European Commission’s High Level AI Experts group, and co-chair of the innovation and commercialization working group of Global Partnership on AI, a global multi-stakeholder initiative which aims to bridge the gap between theory and practice on AI by supporting cutting-edge research and applied activities.

Aleia’s ecosystem of startups includes QWAM Control Intelligence, which specializes in analytics and exploitation of textual data and documents with semantics, AI, and Big Data natural language processing; computer vision expert XXII; Cliris, which uses Machine Learning and Big Data to analyze crowds; Linkurious, which analyzes social graphs; Le Voice Lab, which specializes in analyzing voice; Adobis Group, a specialist in data virtsualization, and Cosmian, which secures Cloud-native apps with advanced cryptography. All of the initial startups are French but Aleia plans to add startups from other European countries.

Startups’ offerings will be integrated into the Aleia marketplace, which is powered by France’s Dawex Data Exchange Platform, a white-label SaaS platform to distribute, source, commercialize and/or orchestrate data ecosystems.

Quam CEO Christian Langevin, a speaker at the launch event, explained how the capacity to process text data at scale is particularly important to the public sector. He said it is difficult for small tech companies to scale due to the risk adverse attitudes of public sector purchasing departments and said this was a driving factor in QWAM’s decision to join the Aleia ecosystem.

Data sharing is another main objective for Aleia. The design of Aleia’s platform will give companies and governments the opportunity to share and exchange data in a controlled manner with the goal of assembling enough data to train algorithms, says Couret.

“The adoption of AI is constrained for many companies, and most notably for the smallest, by a lack of data, the basic material of digital transformation,” he says. “Only by pooling the data of private and/or public organizations can we attain the critical amount needed to accelerate the training of algorithms.”

In addition, “data used by AI is now multi-faceted, with text, images, videos, speech or graph data, so that a one stop shop is needed where one can come and find all the tools to analyze the data and rapidly develop and deploy his or her application,” Soulié says.

The take-up of AI in the public sector for applications like health and the willingness to share public data with private entities is constrained by privacy concerns, says Vedel. This may change later this year after Europe passes the EU Data Governance Act, which is intended to foster the availability of data by increasing trust in data intermediaries and strengthening data sharing across the EU and between sectors. The law will introduce new data governance, in line with EU rules on personal data protection, consumer protection and competition law, as part of the European Strategy for Data.

“We need companies to be confident about exchanging data and not to be afraid of privacy regulationa,” says Couret. “We need to rethink regulation based on a new balance between privacy and efficiency.”

There is another serious issue that needs to be resolved if Europe is going to become an AI leader. Soulié points out that every time a large French or European company awards a contract to an American Cloud provider, European data is being used to improve American AI tools. Most of European data is currently being processed by American tech companies. European AI tools will never be as good or surpass those of the U.S. if they don’t train on massive amounts of data, she says.

The global AI race is largely seen as being between the U.S. and China. Europe is a distant third. If European governments want the region to catch-up large companies need to work more closely with European startups, says Laurent Lafaye, Co-CEO of Dawex, a speaker at the Feb. 3 launch event. He suggested, during a round table, that the French government, which holds stakes in some of France’s biggest companies, should exert pressure on those companies to allocate a portion of their existing R &D budget to working with local AI startups.

Aleia, which currently has around 30 collaborators and expects to triple that number by the end of this year, is actively courting some of France’s largest companies as potential clients.

The pitch is that all data will be hosted on European soil, either with a provider like France’s OVH or within a corporate’s own network, and all processing of data and algorithms will be done using either open source or European software tools, meaning it will not be subjected to the CLOUD Act, which allows federal law enforcement to compel U.S.-based technology companies via warrant or subpoena to provide requested data stored on servers regardless of whether the data are stored in the U.S. or on foreign soil. The offer will include four building blocks: AI Lake, for importing and transforming data; AI Factory to simplify and accelerate the construction of data sets, algorithms, and AI tools; AI Run for deploying, integrating and supervising artificial intelligence applications; and AI Admin for managing the governance of data, APIs and the infrastructure in a sovereign environment.

Big Ambitions

Aleia is well plugged into the French and European AI ecosystems. It is a member of Hub France AI, which regroups the French AI ecosystem. (Couret is president of both Aleia and Hub France AI) It is also a member of the European Alliance on Applied AI and of Gaia-X, a pan-European project that aims to be “a decentralized, secure, transparent digital ecosystem for the European data economy, allowing digital services and data to be shared by any public or private institution without sacrificing data protection and privacy.”

The Cyber Campus will also be part of the Aleia ecosystem. Cybersecurity and AI are intertwined because Europe’s sovereign AI offerings must be secure, says Yann Bonnet, managing director of The Cyber Campus and a former member of the EU’s High Level Expert Group on Artificial Intelligence. Over a hundred entities, including large corporates, public agencies, research organizations, startups and organizations are already involved in the campus. Most participants are French, but the ambition is to connect to other cybersecurity hubs across Europe, Bonnet says. Part of The Cyber Campus’ mission is promoting the sharing of data to reinforce the capacity of public and private entities to handle cyber risks. Among other things, it aims to create a cyber commons studio to share resources and develop sovereign solutions to respond to threats.

Deja Vu

The French state’s support of sovereign tech projects is not new. For example, as part of Project Andromède, which was announced in 2011 as a governmental desire for French-controlled cloud computing, the French state financed two cloud companies: Cloudwatt, which was formed by Orange and Thales, and Numergy, created by SFR and Bull, in the hopes that they could serve as national alternatives to Amazon, Microsoft and Google. The state divested its stakes in 2016. Numergy was absorbed into telecommunications company SFR and Cloudwatt shut down in 2020.

The French government is also one of the investors in Qwant, a French search engine that safeguards data by keeping it on European soil and respecting user privacy. The company, which presents itself as an alternative to Google, is facing some financial difficulties. According to a story in Politico in June of last year Qwant requested an €8 million loan from Huawei, raising concerns that the Chinese telecom equipment vendor, which has been accused by the U.S. government of espionage, could potentially gain visibility or influence on Qwant’s strategy.

Meanwhile, GAIA-X, which was formed at the behest of the German and French governments, has found itself surrounded by controversy. French cloud provider Scaleway announced in November that it would back out of the project and not renew its membership, citing foreign tech company influence as one of the reasons for its departure.

A Different Trajectory

Aleia is on a different trajectory, says Couret. It is a fully private company, with a business and tech focus, and most importantly, is led by an ecosystem of entrepreneurs. He says he is confident that Aleia can make a strong contribution to developing sovereign AI in France and in Europe by promoting data sharing and harnessing the power of Europe’s best AI startups and its ecosystem. “Based on the wealth of deep tech companies in Europe, Europe deserves tohave its own AI offer and it must be created at a European level,” he says.

Visit: https://innovatorawards.org/

The Tech Sector Nearly Destroyed Media. Can It Save It?

After hobbling the news business by grabbing most of the advertising revenue and normalizing the giving away of content for free, Big Tech is using original articles created by journalists at surviving outlets to train its AI models, without giving credit to their work or providing any kind of compensation.

Big Tech companies are, in fact, hoovering up the content not only of newspapers and magazines but artists, authors and musicians, ballooning their own valuations while threatening the livelihoods of content creators. Copyright lawsuits filed against GenAI companies abound, alleging that the way they operate amounts to theft.

Bill Gross, one of Silicon Valley’s most prolific entrepreneurs, believes there is a better way than lawsuits to combat the problem: using tech of his own invention.

Generative AI cannot thrive on a foundation of stolen or uncredited content—it’s neither sustainable nor just, says Gross, CEO of ProRata.ai, a new company that uses tech to enable generative artificial intelligence (GenAI) platforms to attribute and compensate content owners.

Among other things Gross, who has created more than 150 companies with more than 50 IPO’s and acquisitions over the last 30 years, is widely credited with inventing “pay-for-click”, a novel way for search engines to make money on advertising, when he was running a company he founded in 1998 called GoTo.com. Instead of paying for page-views—an old-media model—advertisers pay only when people click on their ads.

Google paid GoTo.com to license its tech and pay-to-click would go on to create a multi-billion-dollar advertising business. Then came along the latest disruption: Generative AI models like ChatGPT that respond to questions with knowledge gained from crawling content without credit or compensation, essentially giving the builders of large language model a free ride on the massive investment made by the small number of surviving media outlets that have built successful business models from online journalism.

At the same time, OpenAI and other AI tech firms — which use a wide variety of online texts, from newspaper articles to poems to screenplays and books – to train chatbots, are attracting billions of dollars in venture capital.

It doesn’t need to be a zero-sum game, says Gross. YouTube, which started out by using other people’s content, saw its business thrive when it started revenue 50/50 with creators, and music streaming service Spotify has paid out billions of dollars to artists so “it is completely possible to pay creators and make a viable business,” says Gross, who spoke about ProRata in January at the DLD technology conference in Munich and at an Axios side event at the World Economic Forum’s annual meeting in Davos.

“Why should Generative AI be an exception?” he asked during an interview with The Innovator.

Whereas Spotify is based on the number of streams, with Generative AI the challenge was to figure out the proportionate contribution to an answer. Gross invented tech that can reverse-engineer where an answer came from and what percentage comes from a particular source so that owners can be paid for the use of their material on a per-use basis, says Gross, who has patented the technology. ProRata pledges to share half the revenue from subscriptions and advertising with its licensing partners, help them track how their content is being used by AIs, and aggressively drive traffic to their websites.

When a user poses a query ProRata’s algorithm compiles an answer from the best information available. At the top of the page there is an attribution bar which specifies where the answer came from. It might say, for example, 30% of this answer came from The Atlantic, 50% from Fortune and 20% from The Guardian. The publications are immediately compensated according to their contribution to the answer and a side panel displays the original articles and enables users to click on the original source to learn more.

Think of it as “attribution-as-a-service.,” says Gross. “Just as Nielsen measures how TV shows are watched to determine what advertisers should pay, we are moderating the output of the queries to determine how much GenAI providers should pay content providers.”

For starters Gross is launching a GenAI search engine called Gist.ai that only consults the archives of participating publishers. Some 400 publishers have signed on so far, including The Atlantic, Time Magazine, Fortune, The Guardian and Skynews, contributing some 50 million documents. So have book authors such as Adam Grant and Walter Isaacson and Universal Music as the same technology can be used to attribute credit to images, music and movies.

Expect other types of content providers to follow. In his presentation at DLD Gross demonstrated how his technology could determine that an image of a masked superhero provided by Meta was generated using 90.3% of material from Marvel Comic images and 6.2% from DC Comics.

Gross plans to charge $20 a month for individuals to use the Pro version of Gist.ai, the same rate charged by ChatGPT. The difference, says Gross, is that Gist.ai will only use trusted sources of information and has the buy-in of the content owners, who are ethically compensated on a per-use basis.

“This empowers the long tail,” says Gross. “You don’t have to be a big brand” to take advantage of the service, he says. A growing number of professional journalists are trying to monetize their content, but many have struggled to make a living using sites such as Substack, which, like Prorata, bills itself as a new economic engine for content providers.

Once more publishers join ProRata it will put pressure on GenAI platforms to share revenue with content providers, says Gross. He hopes to eventually get Microsoft, Amazon and maybe even Google to license its technology. “We want to convince the industry that if you want to crawl people’s content you should share,” he says.

Gross says he was shocked to see statistics from the tech company Cloudflare that demonstrated that 10 years ago Google crawled two pages for every visitor to a website, but today it crawls six pages for every visitor it sends, making it three times harder for content providers to monetize. OpenAI crawls 250 pages for every visitor it sends and Anthropic crawls 250,000 for every one visitor. Why is it so little? “They obscure where the content is from so there is almost no reason to go to a site,” says Gross. “I want us to get to a fairer value exchange. “



He hopes a combination of things will help convince GenAI platforms to compensate content providers. If guilt and lawsuits don’t work if, over time, more and more publishers block their content from being crawled by GenAI platforms the quality of their chatbot’s answers will deteriorate and people will vote with their feet, choosing to instead access answers from the content of trusted publishers, Gross says.

Visit: https://innovatorawards.org/

Why Companies Are Embracing Open Source AI

A survey of more than 700 technology leaders and senior developers across 41 countries found that business leaders are embracing open source AI tools as essential components of their technology stacks. Overall, more than three-quarters of respondents—76%—expect their organizations to increase use of open source AI technologies over the next several years according to the survey, which was conducted by McKinsey, the Mozilla Foundation, and the Patrick J. McGovern Foundation.

The uptake of open source AI by business can be explained by looking at technological and geopolitical trends.

Open source AI innovations are having impact on two key AI technology developments: privacy-centric Edge applications powered by small language models (SLMs) and the emergence of reasoning models with higher inference-time compute, according to the survey report.

Embracing open source is also increasingly part of the political zeitgeist as governments seeks alternatives to U.S. and Chinese closed models. During the global AI Action Summit in Paris in February political leaders expressed concern about concentration of power in the hands of a few AI companies. Some 58 of the countries attending the summit – which together represent one half of the global population – signed a statement committing to promoting AI accessibility to reduce digital divides; ensuring AI is open, inclusive, transparent, ethical, safe, security and trustworthy and avoiding market concentration.
“
“The future of AI belongs to ecosystems, not empires,” says Vilas S. Dhar, president, Patrick J. McGovern Foundation, a contributor to the survey report and a participant at the AI Action Summit.

In a May 1 interview with The Innovator Dhar argued that open source enables open innovation. “By democratizing access to innovation ecosystems, open source puts the tools of creation into everyone’s hands, allowing regionally appropriate AI models to develop,” he says. He points to an initiative by Chile’s Ministry of Science, Technology, Knowledge and Innovation and the National AI Center to launch Latam GPT, a large language model designed to understand and represent the history and culture of Latin America. The official launch is scheduled for June, with capabilities comparable to OpenAI’s ChatGPT 3.5.

Latam GPT was developed with the support of experts, institutions, and research centers across Mexico, Argentina, Colombia, Ecuador, the United States, Spain, Peru, and Uruguay. The project’s vision is to democratize access to advanced AI technologies, ensuring every country in the region can develop and implement AI systems in their governments and industries.

“This takes the idea of open source and uses it to build a product that will drive an entire ecosystem,” says Dhar. “Open source can unlock thousands of new approaches by breaking the link between dependence on costly foundation models and innovation in the fast, affordable applications that are built on top of them.”

Sovereignty through collaboration is an idea that is taking hold in Europe as well. “In a world where exponential technologies have shifted the concentration of power and value capture to a handful of non-European companies, mostly through proprietary and lock-in techniques, openness is the only radical and non-conflicting public policy that can reverse the trend immediately,” Yann Lechelle, CEO of Probabl, a spin-off of French research center Inria that has been financing a global open source data science library called scikit-learn, a tool widely used for performing complex AI and machine learning tasks, said in a recent interview with The Innovator. “For the European Commission stimulating, supporting and adopting more open science, open data, open source and open weights, open standards and open hardware , may be the strongest path to transforming the economic landscape. It’s a weapon that can be used for a massive leveling of the playing field.”

Big Tech is, in fact, playing a big role in open source as well. The most common open source AI tools used by enterprises, as of January 2025, are those developed by large technology players, such as Meta with its Llama family and Google with its Gemma family, according to the survey report.

“I see this as Big Tech recognizing the growing momentum around open source as a path to build more inclusive and participatory ecosystems,” says Dhar. “It also reflects a realization that market dominance will not be about holding on to the AI model but about sustaining and supporting developer communities that build on top,” he says. “More innovation is better for everyone.”

The Advantages of Open Source

Hyperscalers such as Amazon Web Services (AWS), Google Cloud, and Microsoft Azure are releasing industry-specific, cost-efficient SLM models tailored for specialized tasks and distilled into domain-specific tools to power applications for sectors such as manufacturing and finance. But open source developers are also playing an important role in creating these SLMs, enabling the distillation process of general-purpose LLMs with smaller models that can match or even exceed the performance of larger ones, according to the survey report.

Small models enable Edge applications and on-device intelligence for organizations that prioritize latency and/or privacy. Some examples of small-model hubs that distribute open source (and other) models cited in the survey report include the Qualcomm AI Hub, which addresses the needs of Edge AI product OEMs, and Ollama, which offers a framework and tools to deploy open models to the PCs of individual advanced users. The expectation is that hubs like these will add trusted third-party evaluation certification tools, enhancing customer trust and confidence.

The second key trend is the emergence of reasoning models, which employ higher compute during inference time (rather than in their pretraining time) to excel at specific tasks, says the survey report. While the initial wave of reasoning models were proprietary (such as OpenAI’s o1 reasoning model), open source alternatives—including China’s DeepSeek-R1 and a similarly capable model from Alibaba—have quickly followed. Other players are building on and adapting these. The survey report mentions how Perplexity has modified a version of DeepSeek3 to provide more unbiased and accurate information and Smolagents from Hugging Face has created an alternative Deep Research model, challenging offerings from OpenAI and Google DeepMind.

Other open technologies are emerging to help builders optimize and enhance their model-training pipelines and processes. DeepSeek, for example, has continued to offer open source repositories, including parallelism and integration capabilities, for its reasoning models, says the survey report.

While the capabilities of open source models once lagged proprietary ones, base models have improved significantly, says the survey report. And while enterprises may face challenges in tailoring some of the components of reasoning models and the time to value is often longer “the bottom line is that open source offerings now allow model service providers to bring together a full stack of technologies that deliver an effective developer experience, enable modularity, and capture the advantages of community-based development,” says the survey report. The report argues that open source provides organizations greater flexibility and choice to deploy AI either on the Edge or in the Cloud, depending on their privacy, latency, and performance needs. And it says the open source’s operating model and architectural flexibility “can help build more resilient AI systems.”

Navigating The Risks

Amid the benefits and value of open source AI, there are risks, primarily related to security, that could affect their adoption, says the survey report. The most relevant AI risks cited include cybersecurity (62% of respondents), regulatory compliance (54%), and intellectual property (50%).

The survey report recommends four ways businesses can control the risks when implementing an AI model-based system, whether open source or proprietary:Guardrails: The establishment of robust guardrails—such as automated content filtering, input/output validation, and human oversight—can help ensure responsible use and secure outputs.
Third-party evaluations: Conduct regular assessments with standardized benchmarks that allow for certification. During such benchmarking, private evaluations assure that test data sets are kept private from the model.
Documentation and monitoring: Operationally, a software bill of materials can help track version discrepancies and vulnerabilities by maintaining detailed inventories of open source components. Quantitative risk assessments can assess the severity of vulnerabilities in open source systems.
Cybersecurity practices: To secure data privacy and system integrity running models in trusted execution environments may help to ensure sensitive data remains encrypted during processing. Incorporating differential privacy and federated learning techniques during training can prevent models from memorizing confidential information. Strong access controls within model repositories, network segmentation between training and inference servers, continuous monitoring of security incidents, and cryptographic hash verification to confirm that models are from trusted repositories can help address both content safety and cybersecurity challenges in production AI environments, the survey report says.

A Foundation For A More Innovative Future



The survey found that many companies are opting for hybrid open source and proprietary systems. Still, the momentum behind open source AI is undeniable, says Mozilla President Mark Surman, a contributor to the survey report. “In just the past year, we’ve seen countless examples proving that community-driven innovation can not only compete with but even outperform proprietary models,” he said in a statement in the report. “The next big bet is building open tools and a stack that make AI truly accessible—like an AI Lego box that anyone can use. If we get this right, open source AI won’t just be an alternative to closed systems. It will be the foundation for a more competitive, creative, and innovative future.”

Visit: https://innovatorawards.org/

Ant Group Recognized for Innovation Excellence As a Top 100 Global Innovator 2024

HANGZHOU, China--(BUSINESS WIRE)--For the third consecutive year, Ant Group, a global digital technology provider, has been recognized as a Top 100 Global Innovator™ 2024 by Clarivate™, a global leader in connecting people and organizations to intelligence they can trust to transform their world. This recognition reaffirms Ant Group's ongoing commitment to innovation excellence in technologies such as artificial intelligence (AI) and blockchain.

Gordon Samson, President, Intellectual Property, Clarivate, said, “We congratulate Ant Group for being named as a Top 100 Global Innovator again. To feature as a Top 100 Global Innovator is no mean feat as maintaining an edge in the innovation ecosystem is harder than ever. Organizations must balance experimentation and risk with discipline and reward. We measure and rank innovative performance in a dynamic and thorough way, using live thresholds of differentiation. At Clarivate, we think forward by analyzing the quality of ideas, their potency and their impact to identify the world’s top innovators, and this year we reveal the ranking of these innovators for the first time.”

As of the end of 2023, Ant Group had filed 32,459 patent applications globally, with 22,102 patents granted. The top three categories of these patent applications are security technology, blockchain and AI.

"Our recognition by Clarivate as one of the Top 100 Global Innovators is a testament to our innovation capabilities," said Shen PAN, Director of Patents at Ant Group. "We are committed to leveraging technology to build trust and accelerate digital transformation across industries. Our journey is driven by our advancements in key technologies such as blockchain, privacy computing, security technology, Internet of Things (IoT) databases, and most notably, AI.”

In the field of AI, Ant Group has been continuously leveraging its technology to enhance the user experience across its product offerings for years. As of the end of 2023, Ant Group had filed over 3,000 AI-related patents. During Alipay’s 2024 Chinese New Year campaign, AI features in the app attracted 600 million interactions.

To facilitate technological advancement across industries, Ant Group has been making its innovations publicly accessible to developers in the open-source community. For example, Ant Group’s AI infrastructure team open-sourced ATorch, an extension library of PyTorch, that can improve GPU utilization rate up to 60% in large-scale pre-trainings of Large Language Models (LLMs). Meanwhile, the company’s open-sourced Lookahead achieves lossless generation accuracy for LLMs while boosting the inference speeds of LLMs by 2 to 6 times.

By the end of 2023, the number of open-source repositories from Ant Group on platforms such as GitHub had exceeded 1,900. Additionally, according to the 2023 Blue Paper on the Development of Open Source in China published by China OSS Promotion Union (COPU), Ant Group is recognized as one of the top three organizations in the country in terms of open-source contributions and influence.

Leveraging its innovations in technologies such as blockchain, privacy computing, security technology, IoT, and databases, Ant Group provides technology products and services to support the digital transformation and collaboration of global enterprise customers across a variety of industries. These industries include banking, telecommunication, real estate, medicine and energy. The company has garnered recognition from various organizations for its excellence in delivering innovative products and services to customers. For example, AntChain was recognized by Forbes on the Blockchain 50 list for five consecutive years (from 2019 to 2023). In September 2023, ZOLOZ was named as a Representative Vendor for the second consecutive time in the latest Gartner Market Guide for Identity Verification.

Methodology

The Top 100 Global Innovators uses a complete comparative analysis of global invention data to assess the strength of every patented idea, using measures tied directly to their innovative power.

To move from the individual idea strength to identify the organizations that create them more consistently and frequently, Clarivate sets two threshold criteria that potential candidates must meet and then adds a measure of their patented innovation output over the past five years.

For full information on the methodology used to identify the 2024 list, see here.

About Clarivate

Clarivate™ is a leading global provider of transformative intelligence. We offer enriched data, insights & analytics, workflow solutions and expert services in the areas of Academia & Government, Intellectual Property and Life Sciences & Healthcare. For more information, please visit www.clarivate.com

About Ant Group

Ant Group traces its roots back to Alipay, which was established in 2004 to create trust between online sellers and buyers. Over the years, Ant Group has grown to become one of the world's leading open Internet platforms.

Through technological innovation, Ant Group supports its partners in providing inclusive, convenient digital life and digital financial services to consumers and SMEs. In addition, it has been introducing new technologies and products to support the digital transformation of industries and facilitate industrial collaboration. Working together with global partners, the company enables merchants and consumers to make and receive payments and remit around the world.

Visit: https://innovatorawards.org/

'Startup in Shanghai' competition invites international innovators

The 2025 edition of "Startup in Shanghai" International Innovation and Entrepreneurship Competition is now open for global applications, welcoming outstanding projects from innovation-driven teams and enterprises worldwide, the Shanghai Municipal Science and Technology Commission said recently.

This year's application officially launched in late May, and will run through the end of July. International applicants should register through the WeStart TOP100 website.

Applicants must have a core team of at least three members and possess original technologies with clear commercial potential. Projects should fall within key sectors, such as next-generation information technology, biomedicine, high-end equipment manufacturing, new energy, new materials, environmental resources, and new energy vehicles, according to the commission.

The competition will include preliminary, semifinal, and final rounds, with participants pitching their ideas to panels of industry experts and investors. In addition to first, second, and third prizes, a special grand prize will be conferred to top performers. Winners will receive funding, incubation support, investment matchmaking, and opportunities to engage in high-profile innovation events in China.

Upon registering a company in Shanghai, award recipients may access exclusive services offered by designated partner banks. Eligible participants may also be recommended for exposure and engagement opportunities at high-profile activities, including the Pujiang Innovation Forum, Shanghai Science and Technology Festival, China International Import Expo, and China International Industry Fair, according to the commission.

Since its inception in 2012, the competition has attracted over 70,000 startup participants worldwide, offering a premier platform for showcasing cutting-edge technologies and entrepreneurial talent.

Visit: https://innovatorawards.org/