The U.K. government is recasting one of its most visible AI institutions around security, growth and practical deployment. The AI Safety Institute is being renamed the AI Security Institute, while a new MOU with Anthropic points to deeper cooperation between government and foundational AI companies.
From AI safety to AI security
The Department of Science, Industry and Technology announced that the AI Safety Institute will now be called the “AI Security Institute.” The initials and URL remain the same, but the emphasis is changing.
The institute was created a little over a year ago with a different public focus. Its earlier work centered on areas such as existential risk and bias in large language models. Under the new framing, the body will focus more directly on cybersecurity and on “strengthening protections against the risks AI poses to national security and crime.”
That change matters because it signals how the U.K. government wants to talk about AI risk. The concern is not being presented mainly as a broad debate about future harms. It is being tied to institutions, public protection, crime and national security.
Officials are also presenting the shift as part of a wider economic agenda. The government wants AI to help modernize the economy, support industry and accelerate public-sector work. In that context, security becomes a condition for deployment rather than a separate argument against moving quickly.
The Anthropic MOU widens the government’s AI circle
Alongside the renaming, the government announced a new partnership with Anthropic. The agreement does not announce firm services. Instead, the MOU says the parties will “explore” using Anthropic’s AI assistant Claude in public services.
Anthropic will also aim to contribute to work in scientific research and economic modeling. At the AI Security Institute, it will provide tools to evaluate AI capabilities in the context of identifying security risks.
Dario Amodei, Anthropic co-founder and CEO, framed the agreement around public services. “AI has the potential to transform how governments serve their citizens,” he said in a statement. “We look forward to exploring how Anthropic’s AI assistant Claude could help UK government agencies enhance public services, with the goal of discovering new ways to make vital information and services more efficient and accessible to UK residents.”
Anthropic is the only company announced in this particular move, which coincides with a week of AI activities in Munich and Paris. But it is not the only AI company working with the government. A series of tools unveiled in January were powered by OpenAI.
At the time, Peter Kyle, the secretary of state for Technology, said the government planned to work with various foundational AI companies. The Anthropic deal shows that approach extending beyond one provider.
A growth agenda changes the language of risk
The renaming of the institute fits with the government’s AI-heavy Plan for Change in January. In that document, the words “safety,” “harm,” “existential,” and “threat” did not appear at all.
According to the source article, that omission was not an oversight. The government’s priority is to kickstart investment in a more modernized economy, with technology and AI at the center. It wants to work more closely with Big Tech while also building homegrown big techs.
That helps explain why the public message now emphasizes development, AI and more development. Civil servants are expected to get an AI assistant called “Humphrey.” They are also being encouraged to share data and use AI in other areas to speed up how they work.
For consumers, the plan includes digital wallets for government documents and chatbots. These examples show that the government is not treating AI as a narrow research issue. It is positioning AI as infrastructure for public administration and citizen services.
The question is not whether AI safety issues have disappeared. The source makes clear they have not been resolved. The message appears to be that safety concerns should not be handled at the expense of progress.
What officials say will and will not change
Kyle argued that the new name represents continuity as well as a new emphasis. “The changes I’m announcing today represent the logical next step in how we approach responsible AI development – helping us to unleash AI and grow the economy as part of our Plan for Change,” Kyle said in a statement. “The work of the AI Security Institute won’t change, but this renewed focus will ensure our citizens – and those of our allies – are protected from those who would look to use AI against our institutions, democratic values, and way of life.”
Ian Hogarth remains the chair of the institute. He also described the shift as a continuation of work already underway. “The Institute’s focus from the start has been on security and we’ve built a team of scientists focused on evaluating serious risks to the public,” he said. “Our new criminal misuse team and deepening partnership with the national security community mark the next stage of tackling those risks.“
The two statements point to the government’s balancing act. It wants to say that responsible AI development remains important, while making security and economic growth the more visible policy frame.
In practical terms, the renamed institute is being positioned around several linked priorities:
- cybersecurity risks connected to AI;
- national security and crime;
- evaluation of AI capabilities for security risks;
- partnerships with foundational AI companies;
- AI deployment in public services and government work.
A wider shift in AI policy priorities
The U.K. move also lands in a broader political moment. The source article notes that priorities around “AI Safety” appear to have changed further afield.
In the U.S., the biggest risk facing the AI Safety Institute is described as the possibility that it will be dismantled. U.S. Vice President J.D. Vance signaled as much during his speech in Paris earlier this week.
That comparison underscores the direction of travel. AI policy is still concerned with risk, but the center of gravity is moving toward security, deployment, public services and economic growth.
For the U.K., the renamed AI Security Institute is now the clearest symbol of that shift. The institution keeps the same initials, but the political message around it has changed: AI is to be evaluated, protected against and deployed as part of a broader push to modernize government and grow the economy.