Following the recent boardroom power struggle that led to the sacking and reinstatement of OpenAI boss Sam Altman, Microsoft’s relationship with OpenAI is now under US and UK antitrust scrutiny.

What Happened? 

A recent boardroom battle at OpenAI (ChatGPT’s creator and working partner of Microsoft), led to the rapid ousting of OpenAI’s boss Sam Altman and resignation of OpenAI’s co-founder Greg Brockman. Both men were reported to have been immediately hired by Microsoft to launch a new advanced AI research team with Altman as CEO. Then, just days later and following the board being replaced (apart from Adam D’Angelo) by a new initial version, Sam Altman returned and was reinstated as OpenAI’s CEO.

What’s The Issue? 

The factors that appear to have attracted US and UK regulators over antitrust concerns are:

– Microsoft has long been a significant supporter and backer of OpenAI, investing in the company and also integrating OpenAI’s technologies within Microsoft’s own products and cloud services. This collaboration has helped in scaling OpenAI’s research and the implementation of AI technologies, particularly in areas like large language models, cloud computing, and AI ethics and safety. It could also, however, be a kind of background evidence of a close relationship between the two companies.

– As mentioned earlier, when Sam Altman was ousted, Microsoft reportedly immediately hired him as CEO of a new research team there (further evidence of a very close relationship).

– Microsoft has been granted a non-voting, observer position at OpenAI by a new three-member initial board. This means that Microsoft’s representative can attend OpenAI’s board meetings and access confidential information (but can’t vote on matters including electing or choosing directors). However, it’s not yet been reported who from Microsoft will take the non-voting position and what a final (rather than the initial) OpenAI board would look like.

– More specifically, the main concern of regulators appears to be whether the partnership between OpenAI and Microsoft has resulted in an “acquisition of control”. This is whether one party has material influence, de facto control, or more than 50 per cent of the voting rights over another entity. Such control, for example, could negatively impact market competition. The UK’s Competition and Markets Authority (CMA) is particularly looking into whether there have been changes in the governance of OpenAI and the nature of Microsoft’s influence over its affairs.

– The CMA recently stated that it’s considering whether it is (or may be) the case that Microsoft’s partnership with OpenAI (or any changes thereto) has resulted in the creation of a relevant merger situation under the merger provisions of the Enterprise Act 2002. Also, if so, the CMA has stated that it’s interested in whether the creation of that situation may be expected to result in a substantial lessening of competition within any market or markets in the United Kingdom for goods or services. The CMA has opened an investigation of the partnership between Microsoft and OpenAI which is currently at the comments and information gathering stage which closes on 3 January 2024.

– Although OpenAI’s parent is a non-profit company (a type of entity thai is rarely subject to antitrust scrutiny) in 2019, it set up a for-profit subsidiary, in which Microsoft is reported to own a 49 per cent stake. It’s also been reported that Microsoft is prepared to invest more than $10 billion into the startup.

In The US? 

Although the above points relate to the UK, the US Federal Trade Commission (FTC) is also reported to be examining the nature of Microsoft’s investment in ChatGPT maker OpenAI in relation to whether it may violate antitrust laws but hasn’t yet opened a formal investigation.

What Does Microsoft Say? 

Microsoft has stated publicly that it doesn’t own any part of OpenAI. Company spokesman, Frank Shaw, said: “While details of our agreement remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions”. 

Meaning? 

Microsoft’s statement that it doesn’t own any part of OpenAI and is merely entitled to a share of profit distributions addresses only one facet of potential antitrust concerns, i.e. mainly the question of ownership. However, antitrust issues often encompass more than just ownership stakes. They can involve questions of influence, control, or exclusive agreements that might affect market competition.

Regulators may still be interested in the broader implications of the Microsoft-OpenAI relationship. This could include the extent of influence that Microsoft might have over OpenAI’s decisions, the potential for their partnership to impact market dynamics in the AI sector, or any exclusive benefits Microsoft might gain. The focus of antitrust authorities, therefore, often extends to how such partnerships influence market fairness, innovation, and consumer choice.

What Does This Mean For Your Business?

In the aftermath of the boardroom changes at OpenAI, including the dramatic sacking and reinstatement of CEO Sam Altman, the antitrust spotlight has turned to the intricate relationship between Microsoft and OpenAI. This scrutiny, in both the US and UK, may go beyond just speculation of a merger and is likely to look at broader concerns of influence and control within the fast-evolving AI sector. The investigations are, therefore, part of a regulatory interest in ensuring competitive fairness in the fast-growing and evolving AI industry.

For businesses, this could translate into an era of increased oversight on AI collaborations and investments and regulators’ concerns over the concentration of power in the AI industry signals a need for businesses to be cautious. The focus is not just on maintaining competitive markets but also on preventing any monopolistic control over emerging and critical technologies like AI. This evolving regulatory landscape indicates that businesses need to consider the broader implications of their strategic partnerships beyond mere ownership stakes.

Microsoft’s assertion that it doesn’t own any part of OpenAI and is only entitled to profit distributions addresses direct ownership concerns but doesn’t fully alleviate antitrust concerns. The nature of their collaboration, potential influence on business decisions, and any exclusive benefits or access could still be under scrutiny.

The parallel inquiries by the FTC in the US and the CMA also appear to suggest a harmonised approach towards regulating major AI partnerships and means that companies operating transnationally in the AI space must be aware of regulatory developments in multiple jurisdictions. The CMA’s investigation into whether the Microsoft-OpenAI partnership has created a “relevant merger situation” under the Enterprise Act 2002, and its potential impact on market competition, could also set precedents affecting future tech collaborations.

Meta has announced the start of a roll-out of default end-to-end encryption for all personal chats and calls via Messenger and Facebook, with a view to making them more private and secure.

Extra Layer Of Security and Privacy 

Meta says that despite it being an optional feature since 2016, making it the default has “taken years to deliver” but will provide an extra layer of security. Meta highlights the benefits of default end-to-end encryption saying that “messages and calls with friends and family are protected from the moment they leave your device to the moment they reach the receiver’s device” and that “nobody, including Meta, can see what’s sent or said, unless you choose to report a message to us.“  

Default end-to-end encryption will roll-out to Facebook first and then to Instagram later, after the Messenger upgrade is completed.

Not Just Security and Privacy 

Meta is also keen to highlight the other benefits of its new default version of end-to-end encryption for users which include additional functionality, such as the ability to edit messages, higher media quality, and disappearing messages. For example:

– Users can edit messages that may have been sent too soon, or that they’d simply like to change, for up to 15 minutes after the messages have been sent.

– Disappearing messages on Messenger will now last for 24 hours after being sent, and Meta says it’s improving the interface to make it easier to tell when ‘disappearing messages’ is turned on.

– To retain privacy and reduce pressure on users to feel like they need to respond to messages immediately, Meta’s new read receipt control allows users to decide if they want others to see when they’ve read their messages.

When? 

Considering that Facebook Messenger has approximately 1 billion users worldwide, the roll-out could take months.

Why Has It Taken So Long To Introduce? 

Meta says it’s taken so long (7 years) to introduce because its engineers, cryptographers, designers, policy experts and product managers have had to rebuild Messenger features from the ground up using the Signal protocol and Meta’s own Labyrinth protocol.

Also, Meta had intended to introduce default end-to-end encryption back in 2022 but had to delay its launch over concerns that it could prevent Meta detecting child abuse on its platform.

Other Messaging Apps Already Have It 

Other messaging apps that have already introduced default end-to-end encryption include Meta-owned WhatsApp (in 2016), and Signal Foundation’s Signal messaging service which has also been upgraded to guard against future encryption-breaking attacks (as much you realistically can), e.g. quantum computer encryption cracking.

Issues 

There are several issues involved with the introduction of end-to-end encryption in messaging apps. For example:

– Governments have long wanted to force tech companies to introduce ‘back doors’ to their apps using the argument that they need to monitor content for criminal activity and dangerous behaviour, including terrorism, child sexual abuse and grooming, hate speech, criminal gang communications, and more. Unfortunately, creating a ‘back door’ destroys privacy, leaves users open to other risks (e.g. hackers) and reduces trust between users and the app owners.

– Attempted legal pressure has been applied to apps like WhatsApp and Facebook Messenger, such as the UK’s Online Safety Act. The UK government wanted to have the ability to securely scan encrypted messages sent on Signal and WhatsApp as part of the law but has admitted that this can’t happen because the technology to do so doesn’t exist (yet).

There are many compelling arguments for having (default) end-to-end encryption in messaging apps, such as:

– Consumer protection, i.e. it safeguards financial information during online banking and shopping, preventing unauthorised access and misuse.

– Business security, e.g. when used in WhatsApp and VPNs, encryption protects sensitive corporate data, ensuring data privacy and reducing cybercrime risks.

– Safe Communication in conflict zones (as highlighted by Ukraine). For example, encryption can facilitate secure, reliable communication in war-torn areas, aiding in broadcasting appeals, organising relief, combating disinformation, and protecting individuals from surveillance and tracking by hostile forces.

– Ensuring the safety of journalists and activists, particularly in environments with censorship or oppressive regimes, by keeping information channels secure and private.

– However, for most people using Facebook’s Messenger app, encryption is simply more of a general reassurance.

What Does This Mean For Your Business?

For Meta, the roll-out of default end-to-end encryption for Facebook and Instagram has been a bit of a slog and a long time coming. However, its introduction to bring FB Messenger in line with Meta’s popular WhatsApp essentially enhances user privacy and security and helps Facebook to claw its way back a little towards positioning itself as a company that’s a strong(er) advocate for digital safety.

For UK businesses, this move offers enhanced protection for sensitive data and communication, aligning with growing demands for cyber security and providing some peace of mind. However, the move presents further challenges and frustration for law enforcement and the UK government, potentially complicating efforts to monitor criminal activities and enforce regulations like the Online Safety Act. Overall, the initiative could be said to underscore a broader trend towards prioritising user privacy and security in the digital landscape, as well as being another way for tech giants like Meta to compete with other apps like Signal. It’s also a way for Meta to demonstrate that it won’t be forced into bowing to government pressure that could destroy the integrity and competitiveness of its products and negatively affect user trust in its brand (which has taken a battering in recent years).

Following 36 hours of talks, EU officials have finally reached a historic provisional deal on laws to regulate the use of artificial intelligence.

The Artificial Intelligence Act 

The Council presidency and the European Parliament’s negotiators’ provisional agreement relates to the proposal on harmonised rules on artificial intelligence (AI), the so-called artificial intelligence act.

The EU says the main idea behind the rules is to regulate AI based on its capacity to cause harm to society, i.e. following a ‘risk-based’ approach: the higher the risk, the stricter the rules.

Protection & Stimulating Investment 

The comprehensive, world-first draft regulation aims to ensure that AI systems placed on the European market and used in the EU are safe and respect fundamental rights and EU values. The hope is that this will also help stimulate investment and innovation in AI within Europe.

The Key Elements 

Some of the key elements in the draft AI act include:

– Rules relating to high-impact general-purpose AI models that can cause systemic risk in the future, as well as on high-risk AI systems.

– A revised system of governance with some enforcement powers at EU level.

– The extension of a list of prohibitions but with the possibility to use remote biometric identification by law enforcement authorities in public spaces, subject to safeguards.

– Improved protection of rights through the obligation for deployers of high-risk AI systems to conduct a fundamental rights impact assessment prior to putting an AI system into use.

The Key Aspects

The new EU Artificial Intelligence Act covers several key aspects:

– Clarifying the definitions and scope of the proposed act. For example, the definition of an AI system aligns with the Organisation for Economic Co-operation and Development’s (OECD) approach, providing clear criteria to distinguish AI from simpler software. The regulation excludes areas outside EU law, national security, military/defence purposes, and AI used solely for research, innovation, or non-professional reasons.

– The classification of AI systems and prohibited practices. AI systems are classified into high-risk and limited-risk categories. High-risk AI systems must meet certain requirements and obligations for EU market access, while limited-risk ones have lighter transparency obligations. The act bans AI practices considered unacceptable, like cognitive behavioural manipulation and untargeted facial image scraping.

– Any law enforcement exceptions. For example, the draft rules include any specific provisions allowing law enforcement to use AI with safeguards, including emergency deployment of high-risk AI tools and restricted use of real-time remote biometric identification.

– New rules addressing general-purpose AI (GPAI) systems and foundation models, with specific transparency obligations and a stricter regime for high-impact foundation models.

– A new governance architecture. An AI Office within the Commission will oversee advanced AI models, supported by a scientific panel. The AI Board, comprising member states’ representatives, will coordinate and advise, complemented by an advisory forum for stakeholders.

– Penalties. Fines for violations are set as a percentage of the offending company’s global annual turnover or a predetermined amount, with caps for SMEs and startups.

– Rules around transparency and protection of fundamental rights. For example, high-risk AI systems require a fundamental rights impact assessment before market deployment, while increased transparency is mandated, especially for public entities using such systems.

– Measures in support of innovation including AI regulatory sandboxes for real-world testing and specific conditions and safeguards for AI system testing. The act also aims to reduce the administrative burden for smaller companies.

EU Pleased 

The comments of Carme Artigas, Spanish secretary of state for digitalisation and artificial intelligence, highlight how pleased the EU is that it’s managed to be first to at least put a provisional, draft set of regulations together. As she says on the EU’s Council of the EU pages: “This is a historical achievement, and a huge milestone towards the future! Today’s agreement effectively addresses a global challenge in a fast-evolving technological environment on a key area for the future of our societies and economies. And in this endeavour, we managed to keep an extremely delicate balance: boosting innovation and uptake of artificial intelligence across Europe whilst fully respecting the fundamental rights of our citizens.” 

More Than Two Years Away 

However, despite the three days of negotiations and the announced provisional rules it’s understood that the AI act (which they will lead to) won’t apply until two years after it comes into force (with some exceptions for specific provisions). Given that it’s just over a year since ChatGPT was released and that in that short time we’ve also seen the release of OpenAI’s Dall-E,  Microsoft’s Copilot, Google’s Bard and Duet (and now its Gemini AI model), X’s Grok, and Amazon’s Q, you can’t help thinking that effective regulation of AI looks like it will stay some way behind the rapidly advancing and evolving technology for some time yet.

Criticism

The idea of putting the AI act together for the EU got a negative response back in June when it was criticised by 150 executives in an open letter representing many well-known companies including Renault, Heineken, and Airbus. Some of the criticisms included were that the rules are too strict, are ineffective, and could negatively impact competition and opportunity and undermine the EU’s technological ambitions.

What Does This Mean For Your Business?

The provisional agreement on the EU’s Artificial Intelligence Act is a double-edged sword for businesses in the AI sector. On one hand, it establishes a framework for regulating AI technologies, yeton the other, its long gestation period and potential for stringent regulations have raised concerns about its possible impact on innovation and competition for the EU.

The Act’s implementation timeline is actually a crucial factor for businesses. For example, the new regulations won’t come into force until at least two years after being finalised, thereby creating a window of uncertainty. During this period, AI technology will continue to evolve rapidly, most likely outpacing the regulations being put into place. This could all lead to a regulatory framework that is outdated by the time it is implemented, potentially stifling innovation and putting the EU at a technological disadvantage compared to other regions that may have more agile or less restrictive approaches.

Also, the Act’s stringent rules, particularly for high-risk AI systems, could impose significant compliance burdens on businesses. While these measures are intended to ensure safety and ethical use of AI, there is a risk that they might be too restrictive, hampering the ability of European companies to innovate and compete globally. Over-regulation, therefore, could deter investment in the AI sector, hindering the EU’s technological ambitions and possibly leading to a competitive disadvantage in the global AI landscape.

The balance between regulation and innovation is therefore a delicate one. While (what will become) the Act aims to protect fundamental rights and ensure the ethical use of AI, it also needs to foster an environment conducive to technological advancement. If the regulations are perceived as overly burdensome or inflexible, they could inhibit the growth and competitiveness of EU-based AI companies, impacting the broader European technology sector.

The EU’s AI Act may be a significant step towards regulating emerging technologies, but its success will largely depend on its ability to strike the right balance between safeguarding ethical standards and supporting innovation and competitiveness in the AI industry. Businesses must, therefore, prepare for a landscape that could change significantly in the coming years, staying agile and adaptable to navigate these upcoming regulatory challenges effectively.

Google has announced the phased rollout of its new ‘Gemini’ family of large language models with the Ultra version said to rival the abilities of OpenAI’s GPT-4.

What Is Gemini? 

Gemini, which Google describes as its “newest and most capable” large language model (LLM) and representing a “new era” for AI, is a highly advanced and multimodal AI model. Gemini is a foundational model and not a product like a chatbot. This means it’s designed to be integrated into Google’s existing (and future) products such as its Bard chatbot and Google Search.

The Key Difference 

The key difference to competing LLMs is Gemini’s native multimodality, which means it was built from the ground up to understand, process, combine, and generate different types of data seamlessly, i.e. text, code, audio, images, and video.

This approach differs from traditional multimodal models which often train separate components for different modalities and then stitch them together. As a result, Gemini can handle complex tasks involving various inputs more effectively than its predecessors, thereby making it particularly versatile and powerful.

Three Versions 

Google has produced three versions of the Gemini model, each one optimised for specific tasks. These are:

– Gemini Ultra. This is the largest and most capable version of the model, designed for highly complex tasks. For example, Google reports that it excels in various benchmarks, outperforming existing models and even human experts in Massive Multitask Language Understanding (MMLU). Gemini Ultra is particularly strong in fields requiring advanced reasoning and understanding, such as mathematics, physics, history, law, medicine, and ethics.

– Gemini Pro. This model version is versatile and has been optimised for scaling across a broad range of tasks. It has, therefore, now been integrated into Google Bard to enhance its capabilities. This upgrade has reportedly improved Bard’s performance in understanding and summarising information, reasoning, coding, and planning.

– Gemini Nano.  This version is the most efficient model, tailored for on-device tasks. Its efficiency makes it suitable for applications that require AI capabilities directly on mobile devices or other hardware with limited processing power.

Performance 

In terms of performance, Gemini is reported to have shown exceptional results, surpassing state-of-the-art models in many areas. Google claims, for example that Gemini can outperform OpenAI’s GPT-4 platform (which powers ChatGPT) on 30 of the 32 widely-used academic benchmarks!

Gemini has demonstrated advanced capabilities in not just understanding and reasoning across different modalities but also in coding, being able to understand, explain, and generate high-quality code in multiple programming languages.

Adding To Google’s Search? 

As expected, and intended, Google has been reported to be experimenting with integrating Gemini into its Search Generative Experience (SGE), where it has already shown improvements in speed and quality. This integration could have the potential for Gemini to enhance Google’s search capabilities significantly thereby upping the ante in the search engine market.

Downsides? 

Although Gemini’s exceptional abilities point to a “new standard” being set (as described by Gartner’s Chirag Dekate), this kind of power is bound to come with risk and downsides. For example:

– Possible ethical and societal Impacts. AI systems with advanced reasoning capabilities could still make decisions or produce outputs that reflect biases present in their training data, leading to potential ethical issues and unfair representations in sensitive areas.

– Privacy concerns. The extensive data processing capabilities of Gemini could raise significant privacy concerns, especially regarding personal data misuse, and these concerns could increase as this type of model become more integrated into everyday technologies, e.g. Gemini Nano on devices.

– Misinformation and manipulation. Something this powerful and multimodal could have the ability to seamlessly create realistic fake content could be exploited for crime, spreading misinformation, or manipulating public opinion.

– Dependence and skill erosion. A really powerful multimodal model like Gemini could lead to an overreliance on AI which could lead to a decline in human skills and critical thinking abilities.

– Security risks. Powerful AI models like Gemini could become targets for cyberattacks. If compromised, they could be used for malicious purposes, such as generating harmful content or disrupting critical digital infrastructure.

– Economic impacts. The effects of AI-driven automation on employment, job displacement in certain sectors, and inequality are only likely to be increased by Gemini. As already stated, the Ultra version is very strong in areas like mathematics, physics, history, law, and medicine.

– Regulatory and control challenges. The rapid advancement and complexity of AI models like Gemini make it difficult for regulatory frameworks to keep pace.

– Unpredictable outcomes. The increasing complexity of AI LLMs can lead to less transparent and predictable decision-making processes, therefore, making it difficult to understand and manage these systems effectively.

OpenAI Challenger

Sam Altman, OpenAI’s CEO, has indicated that as early as next year, it could be launching its own new ultra-powerful AI products that could compete with Gemini. Open AI also has the backing of Microsoft (which is currently the subject of a CMA antitrust investigation).

What Does This Mean For Your Business?

With the rollout of Google’s Gemini AI, businesses appear to be on the cusp of a new era in AI. Gemini, with its versions Ultra, Pro, and Nano, is not just another large language model, but it represents a leap forward in AI’s ability to understand, process, and generate a multitude of data types, including text, code, audio, images, and video. This multimodal functionality is a key differentiator, setting it apart from existing models in a value-adding way.

For businesses already leveraging Google’s suite of products, the integration of Gemini could mean a significant boost in efficiency and capability. The enhanced Bard chatbot and Google Search, powered by Gemini, are likely to deliver more accurate, nuanced, and comprehensive results. This could transform how businesses handle data, engage with customers, and develop content.

Also, the advanced capabilities of Gemini, especially in its Ultra version, offer unparalleled opportunities in areas requiring deep analysis and reasoning, like market research, product development, and strategic planning. Its ability to outperform other models and even human experts in certain tasks could provide businesses with insights and solutions that were previously unattainable.

However, this power comes with challenges and responsibilities. For example, its power and multimodal capabilities could be effectively exploited by bad actors and the advanced data processing capabilities of Gemini could pose privacy and security risks if not managed carefully. Additionally, as AI technology advances rapidly in this way, staying compliant with evolving regulatory frameworks is crucial and businesses must navigate these changes responsibly to avoid legal and reputational risks. Also, with the EU only just compiling its own provisional AI bill (which won’t become law for at least 2 years), and OpenAI set to introduce its own next generation LLM in 2024 it seems that effective regulation in the AI market looks like being incredibly challenging and likely to lag considerably behind the technology.

The increasing economic impacts of AI-driven automation, particularly in employment, also warrant attention and businesses may be left with decisions such as how to reskill and redeploy their workforce to mitigate the effects of ultra-powerful LLMs and their AI chatbots eating into wider areas of human expertise.

Google’s Gemini, therefore, presents businesses with a wealth of opportunities for growth and innovation and yet, it also underscores the importance of a balanced approach in leveraging AI technology, and the need for regulation to keep up. As the AI landscape continues to evolve, businesses must remain adaptable, ethical, and vigilant to harness the full potential of AI while mitigating its risks. Gemini looks like being a disruptive competitive advantage for Google in the short term. The future competition in the AI market, with companies like OpenAI gearing up to introduce their own advanced models, indicates an exciting and challenging road ahead for businesses navigating the world of AI.

A single 1976 $4.01 cheque from (and signed by) Apple co-founder Steve Jobs has been sold at auction for a whopping $46,063.

The cheque, written just four months after Apple was founded, was written for a purchase at Radio Shack, where co-founder Steve Wozniak famously bought a TRS-80 Micro Computer System.

The components from the computer were used by Wozniak to help build the ‘blue box’ which could enable users to make free long-distance phone calls. Beginning in 1972 the pair reportedly sold around 200 of the blue boxes for $150 each and this was the first commercial collaboration between Jobs and Wozniak.

Steve Jobs is quoted as saying: “If it hadn’t been for the Blue Boxes, there would have been no Apple. I’m 100% sure of that.” 

Toyota Financial Services (TFS), a subsidiary of Toyota Motor Corporation, has warned customers that it recently suffered a data breach which exposed sensitive personal and financial data.
The correspondence with affected customers follows Toyota confirming last month that unauthorised access on some of its Europe (and Africa) systems had been detected. Medusa ransomware reported that it was behind Toyota’s system being compromised and issued Toyota with an $8,000,000 ransom request to have the stolen data deleted.
The advice from TFS to its customers is to contact their bank to take additional security precautions, add 2FA to their online accounts, monitor any unusual activities, and obtain a current credit report from Schufa (a German credit rating agency). Toyota has also said that it has informed the responsible state data protection officer (for North Rhine-Westphalia) in compliance with GDPR.