Several employees at the UK’s AI Safety Institute (AISI) are currently on sick leave and receiving psychological counseling due to work-related stress. According to a report from September 22, 2026, staff face pressure from tight testing schedules and the nature of their findings regarding potential biological and cybersecurity threats within unreleased AI systems.
Internal restructuring in May 2026 saw the societal resilience team merged into the human impacts unit, reducing the combined headcount from approximately 15 researchers to three. Following this change, Andrew Strait, the former head of the societal resilience team, resigned in July 2026. Former employees have described the workplace atmosphere as stressful and toxic.
Established in November 2023 following the AI Safety Summit, the institute focuses on national security risks. Despite its mandate, AISI lacks the statutory authority to force companies to submit models for testing or to prevent the release of AI systems. The UK government has acknowledged these challenges and pledged to improve working conditions for the staff.
