AI accountability comes into focus in UK, US and EU policy

The UK government is consulting until 30 September 2026 on workplace monitoring technologies, including tools for location tracking, biometric access, keystroke monitoring and AI-driven performance evaluation. Proposed principles cover purpose, transparency, worker engagement, fairness, proportionality, human oversight, dignity and accuracy, while existing duties under the UK GDPR and Data Protection Act 2018 continue to apply.

The UK Jurisdiction Taskforce has concluded that English common law can address harm caused by AI systems without immediate legislation, with contract expected to allocate risk across the supply chain and negligence filling gaps. The FCA is also pressing major technology platforms to tackle AI-enabled investment fraud after issuing 2,329 warnings about unauthorised or potentially fraudulent firms in 2025.

California’s AI Transparency Act became operative on 2 August 2026, requiring large generative AI providers to embed disclosures, offer visible labels and provide detection tools for certain synthetic media. The regime runs alongside Article 50 of the EU AI Act, which applies broader transparency duties. Elsewhere, Stanford researchers created 16 functional bacteriophages using generative AI, Big Tech AI capital spending passed $1.1tn, and Nvidia reported $96bn (£71bn) in quarterly revenue.

Public sector deployments are also expanding. Transport for London estimates AI-enabled traffic controls could reduce delays by up to 14% and generate around £1 billion in economic benefits, while UK councils are exploring AI for administrative work, care planning and resident support as they face a projected £3.8bn funding gap by the end of 2027-28.