AI Safety Institute

The AI Safety Institute was announced by the UK government in November 2023 as the evolution of the Frontier AI Taskforce (itself launched in April 2023), becoming, in the government's words, "the first state-backed organisation focused on advanced AI safety for the public interest". Its stated mission was "to minimise surprise to the UK and humanity from rapid and unexpected advances in AI" by building the technical and sociotechnical infrastructure needed to evaluate frontier AI systems, conduct foundational safety research, and act as a trusted intermediary sharing findings across government, industry, academia and the public. The government explicitly stated the Institute was "not a regulator" and would not itself determine AI regulation, consistent with the UK's wider 'pro-innovation' approach of relying on existing regulators rather than a dedicated AI statute.

On 14 February 2025, at the Munich Security Conference, the Technology Secretary announced that the Institute would be renamed: the government "recast the AI Safety Institute the 'AI Security Institute'", a change intended to "reflect its focus on serious AI risks with security implications, such as how the technology can be used to develop chemical and biological weapons, how it can be used to carry out cyber-attacks, and enable crimes such as fraud and child sexual abuse." The renamed body continues the original Institute's evaluation and research functions but with a reoriented emphasis on national security and criminal misuse rather than the broader societal and ethical AI-safety remit originally described in 2023.