According to a Cabinet Office news story published on 23 September 2026, Dr Jade Leung will become Vice-Chair of the AI Security Institute and Security Advisor to the AI Taskforce. The change keeps her inside the government’s AI safety and security structure, but in narrower strategic roles rather than the two full-time posts she currently holds. (gov.uk) The same announcement says she will step back at the end of September from her roles as the Prime Minister’s Adviser on AI and Chief Technology Officer of the AI Security Institute, citing personal circumstances. A replacement adviser to the Prime Minister has not yet been named. (gov.uk)
For policy readers, the main point is that this is not a clean break with government AI work. It is a redistribution of responsibilities: the day-to-day executive posts are being vacated, while Leung remains attached to both the institute and the Prime Minister’s wider AI programme through an advisory brief focused on safety and security. (gov.uk) That matters because the current ministerial portfolio places both the AI Taskforce and the AI Security Institute within the responsibilities of the Minister for Artificial Intelligence. The appointment therefore preserves a line of continuity across the government’s central AI governance arrangements while wider leadership changes are made. (gov.uk)
The AI Security Institute’s own public description presents it as a research organisation intended to build government understanding of advanced AI risks and possible technical responses. Its published work includes testing leading AI systems, informing policymakers and supporting safety research through grants and collaboration. (aisi.gov.uk) In the Cabinet Office account, Leung was one of the senior officials involved in turning that model into an operating institution. Ministers say she helped establish the institute and build its technical research capability and programme of work, which explains why the government has chosen to retain her in a senior strategic position even as she leaves the chief technology officer role. (gov.uk)
The AI Taskforce is the newer part of the picture. When ministers launched it on 24 July 2026, they said it would sit close to the Prime Minister, be chaired by Lord Vallance, report through the Cabinet Secretary and AI Minister Kanishka Narayan, and drive the government’s overall AI strategy across public service reform and economic growth. (gov.uk) Set against that remit, Leung’s appointment as Security Advisor places an AI safety specialist inside the body now responsible for central coordination and delivery. The practical effect is to keep security considerations tied to the government’s wider programme on adoption, growth and state capability. That is an inference from the published remit and the wording of her new post. (gov.uk)
The announcement also sharpens the distinction between operational management and strategic counsel. GOV.UK says Leung will leave the two full-time posts while continuing to provide advice on AI safety and security, and will also take up a part-time position as a Distinguished Visiting Fellow at Stanford University’s Hoover Institution. (gov.uk) For departments, regulators and security officials, the immediate message is continuity on risk policy rather than a reset. The government is signalling that Leung’s expertise will remain available to ministers and the Taskforce even as formal line-management responsibilities move elsewhere. That reading follows from the structure of the new roles and the decision to appoint a successor adviser separately. (gov.uk)
Ministers are presenting the move as a managed transition. In the Cabinet Office release, Prime Minister Andy Burnham said the UK was better prepared for the opportunities and challenges of AI because of Leung’s work, while Narayan credited her with helping to build the institute and contributing to wider government AI initiatives. (gov.uk) That language is politically relevant because it frames the reshuffle as institutional consolidation, not policy retreat. The government is keeping the same senior figure inside its AI safety and security arrangements while it recruits a new adviser to the Prime Minister. That is the clearest indication available from the official announcement. (gov.uk)
In plain terms, the appointment means the UK is trying to preserve experience at the top of its AI safety system while reassigning the most demanding executive posts. The AI Security Institute keeps a visible senior figure linked to its original build-out, and the AI Taskforce gains a named security adviser as it expands its role across Whitehall. (gov.uk) The remaining open question is succession. A new Adviser on AI to the Prime Minister will be appointed in due course, according to GOV.UK, and that choice will determine who takes over the day-to-day link between No.10, the Taskforce and the government’s technical AI institutions. Until then, the official position is continuity in strategy, with change in staffing. (gov.uk)