AI View - July 2026

Welcome to AI View, Simmons & Simmons’ fortnightly round-up of key AI legislative, regulatory, and policy updates from around the world.

AI x Dispute Resolution Webinar Series

Our AI x Dispute Resolution webinar series is now well underway, bringing together lawyers from across our global disputes and AI teams to unpack how AI-related risk is translating into real-world litigation and regulatory exposure.

Across the first sessions, we have explored how disputes are already emerging in practice, from regulatory enforcement and collective actions to negligence, and what organisations should be doing now to prepare.

The series continues after the summer with a focus on some of the most complex and fast-evolving areas of AI risk, including:

• AI and IP - protecting rights and managing risk
• AI incident response and product liability
• Evidence in AI disputes
• Futureproofing against AI disputes through contracting and governance

Each session is designed to be practical and digestible, combining expert insight with clear takeaways for in-house legal, risk and compliance teams navigating this rapidly developing landscape.

If you have not already registered, you can sign up once to attend the remaining sessions and access recordings from those you may have missed here.

This edition brings you:

• Council of the EU gives final approval to AI Act simplification package (“Omnibus VII”)
• US implements ‘No Robot Bosses Act’ targeting AI-only workplace decisions
• US lawmaker introduces AI Incident Reporting Act in House of Representatives
• Bank of England Probes Agentic AI Threats to Financial Stability
• US FTC proposes policy statement addressing AI accuracy
• UN calls for united AI assessment and regulatory framework

1. Council of the EU gives final approval to AI Act simplification package (“Omnibus VII”)

On 29 June 2026, the Council of the EU gave its final approval to a new regulation intended to streamline and simplify certain AI rules as part of the EU’s “Omnibus VII” simplification package. In particular, the regulation:

• Delays the application of the AI Act’s high-risk regime, with the new application dates set at 2 December 2027 for stand-alone high-risk AI systems (Annex III) and 2 August 2028 for high-risk AI systems embedded in products (Annex I).

• Adds a new prohibited AI practice covering the generation of non-consensual sexual and intimate content and AI-generated child sexual abuse material. According to the Council, AI systems that generate nude images of real people or manipulate existing images to reveal intimate parts will be banned from December 2026.

• Postpones the deadline for national competent authorities to establish AI regulatory sandboxes to 2 August 2027.

• Reduces the grace period for providers to implement transparency solutions for AI-generated content under Art. 50(2) from six months to three months, with a new deadline of 2 December 2026 (but only for existing AI systems; not new AI systems).

The regulation further clarifies the AI Office’s supervisory competence in relation to AI systems based on general-purpose AI models where both the model and the system are developed by the same provider, while preserving national competence in certain areas, including law enforcement, border management, judicial authorities and financial institutions. The regulation will now be published in the Official Journal of the European Union and will enter into force shortly thereafter.

Read the Council’s press release here.

2. US implements ‘No Robot Bosses Act’ targeting AI-only workplace decisions

On 18 June 2026, the No Robot Bosses Act (SB 4833, the Bill) was introduced in the US Senate, seeking to prohibit certain uses of automated decision systems by employers. The Bill would apply to employers that use automated decision systems to support employment-related decisions, including hiring, promotion, discipline, scheduling, compensation and termination.

The Bill would establish requirements governing the use of automated decision systems in the workplace. Employers would be prohibited from relying solely on automated decision systems to make employment decisions and would be required to ensure meaningful human review before implementing consequential employment actions. They would also be required to assess such systems for bias, discrimination and other risks affecting workers, and to provide notice regarding their use.

Workers would also be granted rights to access information about automated decision systems used in employment decisions and to appeal decisions through a human review process.

The Bill has been referred to the relevant congressional committees for further consideration.

Read the Bill here (expected to be uploaded soon) and read more here.

3. US lawmaker introduces AI Incident Reporting Act in House of Representatives

On 25 June 2026, US Representative Nathaniel Moran introduced the AI Incident Reporting Act (the Bill), which would establish a federal framework requiring developers of certain high-capability AI models to report dangerous capabilities, security breaches and safety incidents to the Secretary of Commerce. Under the Bill, the Department of Commerce would be required, within 180 days of enactment, to establish capability thresholds for AI models and developers that could pose significant risks to US national security or public safety, and to issue compliance guidelines.

Developers of designated “covered models” would be required to submit a report within seven days of knowing, or reasonably believing, that reportable activity had occurred. Reportable activity would include models attempting to evade human oversight, circumvent safeguards or resist shutdown, unauthorised access to or theft of model weights, capabilities enabling offensive cyber operations against important software or critical infrastructure, autonomous acceleration of more advanced AI development, and capabilities that could materially enable chemical, biological, radiological, nuclear or explosive threats. For incidents presenting an imminent or ongoing risk of serious harm, the Department of Commerce would also be required to notify congressional leadership and relevant committee chairs within 48 hours.

The Bill also provides for supplemental reporting as additional information becomes available, includes protections for sensitive, classified and security-relevant information, and permits information-sharing with other agencies where appropriate. The Secretary of Commerce would be empowered to enforce the regime, including by imposing civil penalties of up to $2 million for violations, with each day of a continuing violation constituting a separate offence.

Read the Bill here.

4. Bank of England Probes Agentic AI Threats to Financial Stability

On 1 July 2026, the Bank of England announced that it is reviewing whether existing regulatory frameworks are capable of addressing the use of agentic AI in financial services, including in payments, trading, cybersecurity and operational functions. Speaking at the European Central Bank Forum, Deputy Governor Sarah Breeden said that current rules were not designed for AI agents capable of acting without direct human instruction, and that requiring human oversight for every action taken by such systems is unlikely to be practical.

Breeden noted that agentic AI systems are already entering financial workflows, including in product recommendations, operational processes and trading-related tasks. She distinguished these systems from more traditional automated tools on the basis that they can pursue objectives and make decisions with less direct human supervision, which may increase the risk of firms’ systems behaving in similar ways where they are trained on similar data or optimised towards similar goals.

Breeden identified cyber resilience as one of the Bank’s foremost financial stability concerns in relation to agentic AI, describing recent advances in AI models for identifying cyber vulnerabilities as a “step change” in capability. She said supervisors need to assess risk across the financial system as a whole, rather than focusing only on individual firms, particularly because the same tools that can strengthen cyber defences may also enable malicious actors to launch attacks at greater scale and speed. She also suggested that recovery planning may need to move beyond isolated outages and account for simultaneous disruption across multiple firms, with options under consideration including stronger recovery requirements for core systems, arrangements to maintain critical services during outages, and the ability to rebuild compromised systems quickly.

Breeden also said regulators are considering market-wide safeguards such as guardrails, circuit breakers and kill switches to limit or halt trading if faulty AI models contribute to serious disruption. More broadly, she indicated that recent developments have exposed gaps in frameworks that the Bank had previously considered sufficient for managing AI-related risks. These concerns align with the Financial Stability Board’s June consultation, which stated that AI agents raise distinct challenges for human oversight and proposed 12 non-binding sound practices covering governance, AI risk management, and cyber, ICT and third-party risks, including the need for firms to define clear roles and responsibilities where AI is used in critical or material functions.

Read Breeden’s speech here.

5. US FTC proposes policy statement addressing AI accuracy

On 1 July 2026, the US Federal Trade Commission (FTC) announced that it is seeking public comment on a proposed policy statement addressing concerns that AI companies may manipulate the behaviour of their AI systems in ways that are contrary to reasonable consumer expectations of objectivity and accuracy.

The proposed statement explains that, under Section 5 of the FTC Act, AI companies could engage in unfair or deceptive conduct where they distort AI outputs to pursue undisclosed ideological objectives, particularly where this conflicts with express or implied representations made to consumers about the effectiveness or suitability of those systems for particular tasks.

The proposed statement also addresses the interaction between federal and state-level AI regulation. In particular, the FTC states that some state laws may be pre-empted to the extent that they conflict with a federal regulatory scheme and cites Colorado’s Artificial Intelligence Act as an example of legislation that appears to coerce companies into altering AI outputs to comply with state ideological objectives.

The proposed policy statement will be published in the Federal Register, with public comments open until 31 July 2026.

Read the proposed statement here.

6. UN calls for united AI assessment and regulatory framework

On 1 July 2026, the UN Independent International Scientific Panel on Artificial Intelligence published its Preliminary Report, which it describes as the first global, independent scientific assessment of AI opportunities, risks and impacts. The report was prepared by a panel of 40 scientists and experts drawn from every region of the world, serving in their personal capacity, and is intended to provide a shared evidence base for global policy discussions ahead of the Panel’s first comprehensive report in 2027. Its findings were due to be presented to governments at the inaugural UN Global Dialogue on AI Governance in Geneva on 6 and 7 July 2026.

The report warns that AI capabilities are advancing more quickly than both scientific understanding and governments’ ability to respond, and that existing safeguards are struggling to keep pace. It identifies a particular challenge for policymakers: effective AI governance depends on robust scientific evidence, but waiting for that evidence to become fully clear may mean that intervention comes too late. The Preliminary Report therefore examines AI across seven domains, including societal applications, economic implications, security and environmental effects, human rights and democracy, child safety, and governance and reliability.

The Preliminary Report also calls for a more joined-up, evidence-based approach to global AI governance, noting that existing international regulation remains fragmented. It argues that legal incentives and regulatory frameworks will be needed to identify and address harmful and fast-evolving AI systems, and emphasises that increasing technical capability does not, by itself, make deployment appropriate.

Read the Preliminary Report here.