TLT's AI Brief: July 2026

United Kingdom

UK Government announces plans to ban Under-16s from social media

The UK Government has announced plans to ban under-16s from social media platforms, positioning Britain alongside Australia as one of the most proactive regulators of children's online lives. Major platforms including TikTok, Instagram, Snapchat, YouTube, Facebook and X would fall within scope, whilst messaging services such as WhatsApp are, for now, exempted.

The intervention goes further than a blanket age-gate. Live-streaming restrictions and bans on stranger communication would extend across gaming sites and other online services. The Government has also turned its attention to AI chatbots: "romantic companion" chatbots designed to simulate intimate relationships or emotional intimacy would face a minimum age of 18, with similar intimate functionalities on AI chatbots more widely restricted for under-18s. This is a notable extension of scope acknowledging that the risks children face online are not confined to conventional social platforms.

However, the regulations are not expected until Spring 2027, so organisations fact a lengthy wait to see exactly what their legal obligations will look like. Enforcement hinges heavily on Ofcom, whose capacity the Government has itself acknowledged is under urgent review. Age assurance technology, meanwhile, remains notoriously difficult to make both effective and privacy preserving. Whether this translates into meaningful protection will depend on implementation.

Read more here: Social media to be banned for under-16s in landmark government move to give kids their childhood back - GOV.UK

ICO launches AI growth lab for the legal sector

On 8 June 2026, the Government announced the launch of an "AI Growth Lab", bringing together regulators (including the ICO, SRA and Legal Services Board) to provide practical guidance to organisations deploying AI systems.

The initiative will initially focus on lawtech and conveyancing, aiming to help firms navigate regulatory requirements while accelerating responsible AI adoption. A key objective is to improve access to justice by enabling faster and more efficient legal services, while maintaining data protection and consumer safeguards.

The initiative reflects the UK's continued sector-led approach to AI regulation, emphasising guidance and cross-regulatory alignment rather than a single AI statute.

Find out more at ICO statement on the Government’s new advisory AI Growth Lab | ICO

New Bill proposes greater transparency for bots accessing UK websites

Damian Hinds MP has introduced a private member's bill, backed by the News Media Association, that would require operators of bots systematically accessing UK websites to declare their identity and the purpose of their activity. Currently, many bots disguise themselves as human traffic or provide false information to evade detection, leaving website owners with no visibility over who is accessing their content or why.

The context is significant. Overall, AI bot traffic rose by 300% in 2025, with agentic AI traffic rising by nearly 8,000%. Bot traffic now exceeds human traffic for the first time, with the vast majority of it unwanted or unverifiable. Without transparency, website owners cannot block harmful bots, negotiate fair access terms, or protect their infrastructure.

The Bill is at its introduction stage in the House of Commons. If it progresses, it would establish a basic but significant transparency standard requiring bot operators to identify themselves.

Read more here: New Bill Targets Anonymous Bots That Target UK Websites - News Media Association

UK Government publishes five futures for AI by 2030

The UK Government's Office for Science has published a major foresight report setting out five possible scenarios for how artificial intelligence could develop between now and 2030. The scenarios are not predictions, each explores a plausible future designed to help UK policymakers and industry explore AI risks and opportunities, and to test strategies for navigating towards a more favourable AI future for the UK's economy, society and public services.

The five scenarios range considerably in character:

• Unpredictable Advanced AI: Highly capable but unpredictable open-source models are released, leading to increased cyber-attacks, misuse by terrorist groups, and unforeseen harms, offset partially by rapid scientific breakthroughs.
• AI Disrupts the Workforce: Capable AI controlled by large tech firms drives productivity gains but also significant job displacement, inequality and public backlash.
• AI "Wild West": A proliferation of moderately capable systems, including those controlled by authoritarian states, fuels widespread disinformation, espionage and intellectual property theft.
• Advanced AI on a Knife Edge: A single highly capable system becomes so powerful it cannot be fully evaluated, raising fears of AI deception and loss of human oversight.
• AI Disappoints: AI improves more slowly than expected, failing to meet investor and public expectations, with benefits concentrated among the technically skilled.

The report reflects a broader recognition that future frontier AI could deliver widespread benefits including productivity gains, better public services, scientific breakthroughs, but that the risks, including enhanced disinformation, cyber-attacks and harmful automated decision-making, could materialise at a far greater scale than today.

Read more here: UK Government Office for Science — AI 2030 Scenarios Report

Europe

EU AI Act: From draft transparency guidelines to final Code of Practice

Two significant developments have occurred in relation to the EU AI Act's transparency obligations under Article 50, which take effect on 2 August 2026.

The European Commission published draft Transparency Guidelines on 8 May 2026, covering how AI-generated content must be marked and labelled. A public consultation on those draft guidelines closed on 3 June 2026, drawing submissions from technology companies, civil society and legal practitioners. Several provisions proved controversial: respondents pushed back on the broad definition of "deep fake", the prescriptive multi-layered marking requirements, and a narrow exception for human editorial review.

The Commission is now reviewing submissions before finalising the guidelines. With the core transparency obligations under Article 50 taking effect on 2 August 2026, there is very little time for businesses to prepare. Whether the Commission will soften the more prescriptive elements in response to industry concern remains to be seen.

In parallel, a voluntary Transparency Code of Practice has been open for adherence, and the draft guidelines made clear that non-signatories can expect greater regulatory scrutiny. That Code of Practice has now been finalised. The European Commission has published the final version of its Code of Practice on Transparency of AI-Generated Content. The Code is a voluntary compliance tool: once formally endorsed by the Commission and the AI Board, signatories can rely on it to demonstrate compliance with Article 50's requirements without needing to prove compliance by other means.

For providers (those who develop or place AI systems on the market), the Code requires machine-readable marking of synthetic content, accessible detection mechanisms and robust quality standards. For deployers (the businesses and organisations that use those systems in their own products or services), it mandates clear disclosure labels, including a standardised EU "AI" icon at the point of first exposure, with some flexibility for artistic or satirical works.

The financial stakes are considerable. Non-compliance with Article 50 can attract fines of up to €15 million or 3% of global annual turnover. That exposure makes the Code's structured framework an appealing safe harbour for many businesses.

Read more here: European Commission — EU AI Act Transparency Guidelines consultation page | Code of Practice on Transparency of AI-Generated Content | Shaping Europe’s digital future

European Commission proposes Cloud and AI Development Act

The European Commission adopted a proposal for the Cloud and AI Development Act (CADA), published on 3 June 2026, with the aim of strengthening the EU's cloud and AI ecosystem, investment and infrastructure. It focuses on three core objectives:

• Research, development and innovation: supporting the development of next-generation cloud and AI technologies, including frontier AI, industrial AI and physical AI, with "grand challenges" to drive future R&D efforts.
• Capacity: at least tripling the EU's data centre capacity within the next five to seven years, simplifying permitting and improving access to energy, land, water and financing.
• Autonomy: introducing a single EU-wide sovereignty framework to assess cloud and AI sovereignty, accelerating cloud and AI rollout for critical sectors, and establishing a common procurement framework for public administrations.

The CADA also defines cloud and AI sovereignty across four assurance levels, ranging from Level 1 (data processed and stored within the EU) through to Level 4 (full transparency and control over the software supply chain, with no third-country interference). The proposal reflects the EU's growing concern about digital dependence.

Find out more here: Proposal for the Cloud and AI Development Act (CADA) | Shaping Europe’s digital future

United States

Growing enforcement pressure on OpenAI

A coalition of state attorneys general (reportedly 42 in total) has launched a coordinated investigation into OpenAI, focusing on consumer protection, data use, and risks to minors. This follows Florida's earlier lawsuit and signals a shift towards multi-state enforcement action rather than isolated claims.

Florida has filed a landmark lawsuit against OpenAI and CEO Sam Altman, alleging the company knowingly released addictive and unsafe chatbot technology (please see last month’s AI Brief ). The complaint seeks restrictions on data collection from under-13s. OpenAI rejects the claims and emphasises its safety measures and protections for minors.

Courts are also seeing a rapid increase in litigation filings assisted by the use of AI, with evidence suggesting AI is contributing to a surge in claims brought by self-represented litigants.

Taken together, these developments suggest that generative AI is now firmly within the scope of consumer protection, product liability and procedural law debates in the US.

Read more here: OpenAI under investigation by group of state attorneys general, source says | Reuters

Global

UAE establishes new AI and data authority

The UAE government has approved the creation of a dedicated Artificial Intelligence and Data Authority, merging three existing bodies into a single national entity. Approved by the Prime Minister on 14 June 2026, the authority will oversee the UAE's national AI strategy, unify data platforms, set standards for AI and digital governance, and drive the digital economy's contribution to GDP.

Read more here: UAE creates dedicated AI and data authority to 'build government of the future'

Singapore issues proposed guidelines on personal data in generative AI

Singapore's Personal Data Protection Commission (PDPC) has issued proposed guidelines clarifying how existing data protection law applies to generative AI. The guidelines address the full AI lifecycle, tackling  issues including web scraping, consent for user-provided data, and transparency obligations around model training.

Perhaps the most practically relevant update is the guidance on the “publicly available exception”. Organisations may scrape personal data without consent where it is genuinely and freely accessible to the public, though data behind paywalls or registration walls will need to be assessed on a case-by-case basis.

Furthermore, the PDPC signals that vague references to "product improvement" will no longer suffice as consent for AI training purposes, requiring specific, meaningful notification.

Yet the guidelines are non-binding, and their practical force remains unclear. The real test will be whether the final guidance carries sufficient weight to shift industry behaviour.

Read more here: PDPC | Public Consultation on the Proposed Advisory Guidelines on Use of Personal Data in Generative AI

SpaceX completes landmark IPO

SpaceX has completed what is expected to be the largest IPO in history, raising around $75 billion at a valuation now over $2 trillion. The offering priced at $135 per share on 15 June 2026, with the valuation rising sharply in the days following.

Read more here:  SpaceX shares gain for second day after blockbuster debut

OpenAI and Anthropic file for IPOs

OpenAI and Anthropic have both filed for IPOs, targeting valuations up to $1 trillion and $965 billion respectively. The listings mark a shift as leading AI developers move into public markets, putting the AI sector under sustained investor scrutiny for the first time.

Read more here: OpenAI spending hit $34bn last year ahead of planned IPO

OpenAI delays wider release of GPT-5.6 and ChatGPT "Superapp" overhaul reported

OpenAI has delayed the wider release of its GPT-5.6 model at the request of the US government, initially limiting access to a small group of approved partners while a vetting framework is developed. The delay has drawn pushback from companies concerned about restricted access and is a notable sign of governments taking a more active role in managing the rollout of advanced AI.

OpenAI is separately reported to be planning a major redesign of ChatGPT into a "superapp" integrating coding tools, AI agents and partner services. The overhaul is said to be driven by the need to grow revenue ahead of the potential IPO.

Read more here: OpenAI staggers AI model release after Trump administration request | OpenAI | The Guardian | OpenAI plots biggest ChatGPT overhaul since launch

Half of London businesses report AI skills gap

Half of London businesses report growing skills gaps as AI adoption accelerates, with demand rising for advanced digital, critical thinking and decision-making capabilities. While most businesses are increasing investment in training, a survey of more than 2,000 business leaders found that many still struggle to keep pace with rapidly evolving skills requirements.

Read more here: Half of London firms report skills gap amid AI boom - BBC News

London Tech Week concludes with £6bn in AI investment commitments

London Tech Week 2026 concluded with over £6 billion in AI investment commitments and thousands of new jobs, reflecting the UK's push to position itself as a global AI leader. Major US and international firms announced expansion plans alongside new government-backed infrastructure funding.

Read more here: Britain powers ahead on AI with billions of pounds of new investment and thousands of jobs secured as London Tech Week wraps up - GOV.UK

AI tools piloted in the Crown Court

The UK Government is piloting AI tools in the Crown Court to speed up justice and reduce backlogs. New AI assistants will support tasks such as legal research and case analysis, while tools for judges will help prioritise and group trial-ready cases. Additional AI systems (such as transcription tools for probation officers and tribunals) aim to cut administrative work and free up staff time. All technologies will be tested in controlled environments to ensure safe use, forming part of a wider push to modernise the justice system.

The Law Society of England and Wales has welcomed the move but notes that "while new technology should enhance access to justice, it cannot replace vital funding and additional court staff. Robust safeguards are needed to protect us all and preserve the integrity of the justice system."

Read more here: AI tech ambition to deliver smarter justice for victims - GOV.UK

AI unlocking hidden galaxies from archive data

Astronomers are using lightweight AI tools to make discoveries from both historic and incoming space data. Researchers at the European Space Agency used an AI tool to examine decades of archive images taken by the Hubble Space Telescope, identifying hundreds of previously overlooked galaxies and unusual cosmic structures.

At Oxford, an AI tool is being used to sift through data from supernovae (the explosions that occur when stars die), cutting the amount of manual work required by 85%.

At Birmingham, scientists are using AI to run complex calculations about how stars form and change over time. Tasks that would previously have taken months to process can now be done in milliseconds.

Importantly, many of these AI tools do not require vast computing resources to run. The idea that AI must be expensive and energy-hungry is being challenged by the science community.

Read more here: How AI and an astronomer’s laptop can bring new galaxies within reach

7 -10 July 2026 - UN: AI for Good Global Summit (Geneva)

23 - 24 September 2026 - Big Data London