California and Anthropic: Claude Half Price for Government

An Innovative Agreement Between California and Anthropic

California has recently entered into a groundbreaking agreement with the company Anthropic, aimed at integrating artificial intelligence into its public services. This agreement allows the state to access Claude, Anthropic's AI assistant, at a reduced rate of 50%. The goal is to modernize public services while using AI as a decision-support tool, without replacing the work of agents. This initiative could serve as a model for other administrations facing the high costs of AI technologies.

California Governor Gavin Newsom finalized this partnership, which offers all state agencies, as well as cities and counties, access to Claude at a preferential rate. In addition to the price reduction, the agreement includes training, technical support, and guidance to encourage responsible AI adoption. This strategy aims to improve the efficiency of public services without replacing human agents.

Claude, an AI Assistant for the Public Sector

The agreement between California and Anthropic stipulates that Claude will become the go-to AI assistant for all public agencies via the SITeS (Statewide Information Technology Shared Services) portal. Administrations will benefit from this solution with a 50% discount, while teams will receive free training and specialized technical support.

Claude will be used to assist agents with repetitive and time-consuming tasks, such as document drafting, information analysis, report preparation, and facilitating daily administrative tasks. However, Governor Gavin Newsom has made it clear that AI should not replace public officials. According to him, these technologies should enable agents to work more efficiently and enhance the services provided to citizens.

This approach aligns with the decrees published by California since 2023 to regulate the adoption of generative AI in public administrations. The March 2026 decree strengthened data protection and civil rights requirements while encouraging the deployment of AI when it improves public services.

Anthropic and Its Relations with the Federal Government

This partnership with California allows Anthropic to strengthen its ties with its home state, despite strained relations with the U.S. federal government. Earlier this year, Anthropic opposed the U.S. Department of Defense over the terms of use for Claude in a government contract. The company sought to include guarantees preventing the use of its technology for citizen surveillance or the deployment of autonomous weapons without human oversight.

Since these conditions were not accepted, the Pentagon ultimately reached an agreement with OpenAI. Anthropic was even classified as presenting a risk to the supply chain, limiting its participation in certain defense-related contracts. However, California chose not to address these issues during negotiations with Anthropic, prioritizing a logic of modernizing public services.

Chris Given, the state's Chief Information Officer, indicated that the issue of supply chain risks simply was not discussed during negotiations with Anthropic. This situation illustrates the fragmentation of American policy on artificial intelligence, with some states developing their own strategies independently of Washington.

The Strategic Impact of the Agreement on the AI Market

Beyond its local impact, this agreement demonstrates that price is a crucial lever for the adoption of AI assistants in large organizations. A 50% reduction in cost changes the economic equation, especially when professional subscriptions need to be deployed at scale. Anthropic offers a comprehensive package that combines licenses, support, training, and industry expertise.

California is not starting from scratch, as Claude is already being used in several public projects. The assistant contributes to the development of Engaged California, a citizen engagement platform, and the Poppy project, a tool designed to simplify certain administrative tasks. Claude is also involved in cybersecurity projects, with Claude Security and Claude Code serving to analyze and correct computer code.

Some large administrations are already using the tool in their daily operations. The California DMV relies on Claude to improve customer service and reduce wait times, while the Department of Health Services uses it to optimize internal processes related to Medicaid.

For Anthropic, this widespread adoption across public agencies represents an important showcase, as California is a global hub for AI innovation. A successful deployment could serve as a reference for other U.S. states and large organizations looking to integrate generative AI in a regulated manner.

California is thus betting that responsible AI adoption can enhance the efficiency of public action without compromising the central role of agents, while combining cost reduction, strengthened governance, and operational support.