Anthropic

Defense Secretary Pete Hegseth has given AI firm Anthropic a Friday deadline to grant broad military access to its models or face severe repercussions, potentially including designation as a “supply chain risk” or invocation of the Defense Production Act. This ultimatum arises from a dispute over AI usage, with Anthropic seeking ethical assurances against autonomous weapons and mass surveillance, while the DoD demands agreement to “all lawful use cases.” Despite a significant DoD contract, Anthropic has other major commercial clients and substantial funding.

Anthropic launches Claude Cowork, an enterprise AI tool designed to boost knowledge worker productivity. The platform integrates with business tools like Google Drive and DocuSign, and offers customizable plugins for specific industries. Anthropic aims to make AI more accessible across professional roles, while acknowledging potential disruption for certain job functions. This move intensifies competition in the lucrative enterprise AI market.

Software stocks rallied Tuesday, boosted by Anthropic’s enterprise agent announcements. The AI startup’s Claude Cowork platform updates and new partnerships suggest AI integration rather than displacement. This eased investor fears, leading to gains in Salesforce, DocuSign, and Thomson Reuters. Analysts believe AI’s competitive threat to software may be overblown, emphasizing AI’s reliance on existing data and workflows. This shift in sentiment led to a broad market recovery, including IBM, which had previously seen declines.

U.S. AI firm Anthropic accuses Chinese companies DeepSeek, Moonshot AI, and MiniMax of coordinated “distillation attacks.” The firms allegedly used tens of thousands of fake accounts to extract proprietary information from Anthropic’s Claude model, aiming to quickly replicate its capabilities. This follows similar accusations against DeepSeek by OpenAI, raising national security concerns over potential misuse of advanced AI by authoritarian regimes.

Anthropic CEO Dario Amodei will meet with Defense Secretary Pete Hegseth to discuss integrating Anthropic’s AI into the U.S. military. A key sticking point is Anthropic’s demand that its AI not be used for autonomous weapons or surveillance of citizens, while the DOD seeks broad lawful applications. Despite past disagreements, Anthropic has successfully deployed its models on classified networks and holds a significant DOD contract. This meeting is crucial for navigating their complex relationship.

IBM shares dropped 11% following Anthropic’s announcement of Claude Code’s new AI capability for automating COBOL modernization. This development highlights the disruption AI poses to legacy systems, which power critical sectors like finance. Anthropic’s AI can now streamline complex code analysis, documentation, and risk identification, potentially reducing modernization costs. This event reflects broader AI-driven market anxieties, impacting tech stocks, and signals growing investor concern over AI’s implications for companies reliant on older technologies.

At the India AI Impact Summit, OpenAI’s Sam Altman and Anthropic’s Dario Amodei eschewed a handshake for a fist bump during a group photo, highlighting their intense rivalry. This subtle gesture underscored the fierce competition between the two AI giants, vying for global dominance. Both companies are racing to develop advanced AI models and secure market share, while also publicly debating AI safety and advertising strategies. The event showcased both the competitive spirit and the acknowledged need for collaboration in shaping AI’s future.

The Pentagon is reviewing its relationship with AI firm Anthropic due to disagreements over its model usage. While Anthropic seeks to restrict its AI from autonomous weapons and citizen surveillance, the DOD wants unrestricted use for all lawful purposes. This dispute could lead to Anthropic being classified as a supply chain risk, impacting its contracts and standing, despite recent substantial funding.

Anthropic has launched Claude Sonnet 4.6, their latest LLM, just weeks after releasing Opus 4.6. This new model significantly boosts capabilities in programming, design, and data processing, becoming the default for free and Pro users. Sonnet 4.6 offers “much-improved coding skills” and performance previously seen only in Opus-class models, intensifying the AI development race. This rapid advancement, coupled with substantial funding, positions Anthropic as a major competitor in the AI landscape.

Anthropic has appointed Chris Liddell, a former White House official and executive, to its board. This strategic move aims to help the AI startup navigate complex regulatory and political landscapes surrounding generative AI. Liddell’s extensive experience in government and corporate governance is expected to be invaluable as Anthropic, known for its Claude AI models, seeks to balance technological advancement with responsible development and engage with policymakers.