AI Firm Challenges Pentagon Over Trump Administrations AI Policies

US Defense Department Demands Anthropic Relax Rules on AI Tool Use

The US Defense Department has issued an ultimatum to Anthropic, a leading artificial intelligence developer, requiring the company to loosen its rules governing the use of its AI tools by the Pentagon. The demand comes after reports emerged that Anthropic’s Claude software was used in a classified military operation resulting in the abduction of Venezuelan President Nicholas Maduro in January.

According to sources, US Defense Secretary Pete Hegseth has given Anthropic until Friday to revise its guidelines, or risk losing its government contract worth up to $200m. However, Anthropic is refusing to compromise on its safeguards, which prevent its technology from being used for domestic surveillance and autonomous weapons that can operate without human intervention.

Anthropic was the first AI developer to be used in classified operations by the US Defense Department, established at the Pentagon in Washington, DC. The company’s Claude software is a large language model (LLM) capable of generating text, visual or audio output similar to human content after analyzing massive datasets.

The Pentagon has awarded defence contracts to four AI companies, including Anthropic, Google, OpenAI and xAI, each worth up to $200m. Anthropic was also the first AI company to be approved for classified military networks, working with partners such as Palantir Technologies.

A recent resignation by an AI safety researcher at Anthropic has highlighted concerns about the use of AI in sensitive applications. The researcher cited “interconnected crises” and the difficulty in letting values govern actions within organizations.

The Pentagon’s vision for military AI systems emphasizes their operation without ideological constraints, with Secretary Hegseth stating that AI will not be subject to “woke” ideologies.

Source: Al Jazeera