The British Ministry of Sciences, Technology and Innovation (DSIT) has signed a memorandum of understanding with Anthropic, an American pioneer company in the development of advanced AIs, supported by Google and Amazon. This agreement should allow the United Kingdom to position itself among the world leaders of AI applied to public services. “” “The partnership concluded with Anthropic concretizes our commitment to the United Kingdom the best place to develop the avant-garde AI, so that the advantages of this transformative technology are felt by people and businesses throughout the country ”, said Peter Kyle, British Secretary of State for Technology.
The agreement revolves around several axes. It relates to the integration of “Claude”, a large model of language, developed in 2023, to strengthen the automation of administrative tasks, streamline users’ access to information and make public services more intuitive and interactive.
Archaic practices
Prime Minister Keir Starmer, who is committed to making the United Kingdom a “superpower” of artificial intelligence, has multiplied initiatives to stimulate the productivity of public services thanks to AI. In October 2024, the government had concluded an agreement with Microsoft which enabled British public organizations to access the most powerful AI tools. At the beginning of this year, Keir Starmer announced its intention to deploy AI more widely in public services in order to end archaic practices and which make the state lose more than 50 billion euros per year. Anthropic Claude is already used by many institutions around the world. The European Parliament has adopted it to accelerate documentary analysis and reduce the 80 %search time.
-In the United States, the city of Washington DC deployed a bilingual IA chatbot, based on Claude, to improve accessibility to public health information and strengthen the engagement of residents while US intelligence agencies integrate Claude into their secure environments to improve data analysis and decision-making processes. In the United Kingdom, the Swindon municipal council (southwest of England) has ahead of the government by launching Simply Readable, a solution exploiting the robot to make administrative documents more accessible to people with cognitive disorders.
Evaluate safety risks
In a press release, Anthropic stressed that the potential of its AI model will improve the way in which British citizens access and interact with online information and government services. The company added that the agreement with the British government will also allow “To establish best practices for the responsible deployment of AD A AC capacities in the public sector”. Anthropic plans in particular to work with the British IA security institute to study and assess the capacities of this technology and its risks in terms of security.
“Together, we are committed to developing solid protection measures that will contribute to exploiting the important opportunities presented by AI while guaranteeing its responsible and secure deployment”assured the company. She also promised, within the framework of her agreement with the United Kingdom, to support research and innovation in the country by stimulating the local ecosystem of start-ups and universities specializing in AI. Finally, it will set up the Economic index of Anthropic, a tool analyzing the impact of AI on the labor market which will provide key data to the government to adapt its employment and innovation strategies.
