UK AI Security Institute opens applications for its Alignment Project
The £15 million initiative pools UK, Canadian and Australian government money with funding from Anthropic, OpenAI, Microsoft and AWS, plus compute credits.
- Safety & alignment
- Government & policy
- Colour
The UK AI Security Institute opened applications for the Alignment Project, a £15 million ($20 million) international fund for AI alignment, control and interpretability research. It pooled money from the UK and Canadian AI safety institutes and Australia’s AI Safety Institute alongside philanthropic and industry backers including Schmidt Sciences, Anthropic, OpenAI, Microsoft and AWS, and offered grants of up to £1 million per project plus dedicated compute credits and venture-capital funding for commercial alignment work. Its advisory board included Turing Award winners Yoshua Bengio and Shafi Goldwasser.
UK technology secretary Peter Kyle framed the goal as protecting “the British public” from serious AI risks as capabilities advanced. The first funding round drew over 800 applications from 466 institutions across 42 countries, and later awarded more than £27 million across 60-plus projects — exceeding the fund’s initial £15 million commitment.