Anthropic’s AI product Claude recently rolled out a Word document plugin in beta, supporting drafting, editing, and revisions while automatically preserving the original formatting. The plugin can share context with Excel and PowerPoint, enabling cross-document collaboration. At this stage, it is only available to Team and Enterprise users.

Gate News message: On April 11, Wall Street banks such as Goldman Sachs and Morgan Stanley are testing Anthropic’s Mythos large language model.

Anthropic and CoreWeave have reached a multi-year agreement to rent its data center capacity to support growing demand for AI services. The deal will help Anthropic build its Claude AI models and includes multiple Nvidia chip architectures. The partnership will also require working with Broadcom and Google to secure 3.5 gigawatts of power.

Gate News update: On April 10, Uniswap founder Hayden Adams posted on a social platform today, asking whether there are any Anthropic (an AI technology company) related resources for coordination, and said he hopes to test Mythos on Uniswap.

Musk confirmed on the X platform that Grok 4.2 has 0.5 trillion parameters. Calculations suggest that Claude Opus has about 5 trillion parameters, and Sonnet has 1 trillion. Although Anthropic has not disclosed the model parameters, these figures come from Musk’s remarks. Meanwhile, SpaceX AI’s Colossus 2 supercomputer is training a model with the largest scale of 10 trillion parameters, which will become a key competitive factor.

The U.S. CIA Deputy Director Michael Ellis said the CIA will incorporate AI into its analytical platforms over the next few years, but humans will still be the decision-makers. He mentioned a dispute with Anthropic and emphasized that competition between the U.S. and China in technology innovation is intensifying. AI and blockchain analytics will be key areas.

AI company Anthropic is considering designing its own AI chips to address the chip shortage. Although the plan is still in its early stages, it may ultimately choose to buy off-the-shelf chips; at present, there is no specific design plan or dedicated team.

The U.S. Court of Appeals for the District of Columbia Circuit rejected Anthropic’s motion to pause the Department of Defense from listing it as a security risk, saying that military interests take priority over Anthropic’s harms. The dispute stems from a breakdown in its contract negotiations with the Pentagon, with Anthropic insisting that its AI technology not be used in lethal weapons. The court acknowledged that Anthropic may suffer harm, but said the case should be expedited.

Artificial intelligence company Anthropic has launched the “Glasswing project,” using AI model Claude to detect and patch software vulnerabilities, improving cybersecurity. The initiative brings together multiple companies for coordinated defense, advancing the use of AI in security and potentially reshaping industry standards. Successful deployment will enhance vulnerability detection efficiency and user trust, becoming a core tool for enterprises to safeguard the security of their systems.

AI company Anthropic has restricted access to its new AI system, Claude Mythos Preview, after finding thousands of critical software vulnerabilities in testing. Many of these vulnerabilities had gone unnoticed for years, including security issues across multiple operating systems and cryptographic protocols. Industry experts warn that AI-driven cyberattacks will grow significantly, and that the balance between technical innovation and security defenses must be prioritized.