The United Kingdom drops out of the “safety” by its AI organ, now called AI Security Institute, Inks Mou with Anthropic

The UK government wants to make a difficult turn to raise its economy and industry with AI, and as part of it, it is aimed at an institution that founded just over a year ago for a completely different purpose. Today, the Ministry of Science, Industry and Technology has announced that it will rename AI’s AI Safety Institute of AI. In doing so, it will be transferred from a basic study of areas such as existential risk and bias in large linguistic models, to focus on cybersecurity, more special “strengthening protection against the risks that AI represents national security and crime.”

In addition, the government has announced a new partnership with anthropica. No solid services, but Mou states that the two will “explore” using an anthropic Claude assistant in public services; And the anthrop will aim to contribute to work in research and economic modeling. And at the AI ​​Security Institute, he will provide an AI assessment tools in the context of identifying security risks.

“AI has the potential to transform how governments serve their citizens,” the co -founder and CEO of Anthrop said in a statement. “We look forward to exploring how AI’s AI assistant Claude could help government agencies of the UK to improve public services in order to find new ways to make vital information and services more effective and accessible to the UK residents . “

Anthropic is the only company that is announced today – it coincides with the week of AI activities in Munich and Paris – but this is not the only one working with the government. A series of new instruments that were discovered in January were powered by Openai. (At that time Peter Kyle, the Secretary of State for Technology, said the government plans to work with various founding AI companies, and that is what the anthropic transaction is proven.)

AI-AI Safety Institute’s Government Switching A little more With many fanfare – AI’s security should not be too surprised.

When the newly installed Labor Government announced its AI-T-Treasury Plan to change in JanuaryIt was noticeable that the words “safety”, “harm”, “existential” and “threat” did not appear in the document at all.

It was not supervision. The government’s plan is to launch investments in a more modernized economy using technology and more special AI for this. He wants to work more with Big Tech and also wants to build his own homemade technologies. The main messages it promotes have development, AI and more development. Civil servants will have their own AI assistant called “Humphrey“And they are encouraged to share data and to use AI in other areas to accelerate how they work. Users will be Obtaining digital wallets for their government documents and chatbots.

So, AI safety problems have been solved? Not exactly, but the message seems to be that they cannot be considered at the expense of progress.

The government claims that despite the name change, the song will remain the same.

“The changes I announce today represent the logical next step in how we approach the responsible development of AI – it helps us to unleash AI and grow the economy as part of our change plan,” Kyle says in a statement. “The work of the AI ​​Security Institute will not change, but this new focus will guarantee that our citizens – and those of our allies – are protected by those who would like to use AI against our institutions, democratic values ​​and way of Life. “

“The institute’s focus from the outset was on security, and we have built a team of scientists focused on evaluating serious public risks,” added Ian Hogart, who remains chairman of the institute. “Our new team for criminal abuse and deepening partnership with the National Security Community note the next stage of dealing with these risks.”

Further, it seems that priorities have definitely changed around the importance of “AI safety”. The biggest risk that AI’s AI Institute is considering at the moment is that it will be disassembledS Vice president JD VANCE They telegraphed so early this week during their speech in Paris.