Anthropic

On Wednesday, the AISI, which evaluates AI models for the British government, said both Anthropic’s Claude Mythos Preview and OpenAI’s GPT-5.5 showed progress well above previous trends on cybersecurity testing. Separately, XBOW released data suggesting “frontier models have taken a major step forward in vulnerability discovery.”

It’s a package of “connectors,” installed via a toggle switch, that allows Claude to work inside tools like Intuit Quickbooks, PayPal, Docusign, HugSpot, Google Workspace, and Microsoft 365. “It can plan payroll, close the month, run a sales campaign, chase invoices, and more,” per Anthropic’s blog post.

As the AI adoption in the legal field ramps up, Anthropic has announced that its AI chatbot can now connect to many of the apps used by lawyers, including DocuSign, Box, Thomson Reuters, Harvey, and more. “Claude reviews contracts, surfaces case law, and drafts across the tools your team already uses,” according to Anthropic.

The AI startup says “dreaming” will allow Claude to review previous sessions to “find patterns and help agents self-improve.” This feature, which is rolling out in research preview, is supposed to help AI agents identify frequent mistakes, spot tasks they might converge on, and understand a team’s preferences.

Anthropic is doubling five-hour rate limits for many Claude Code users, removing Claude Code’s peak hours limit reduction, and significantly increasing API rate limits for Claude Opus models, starting today.

Asked by Senator Jacky Rosen (D-Nevada) in a Senate Armed Services Committee hearing about the DoD’s dispute with Anthropic and whether he could guarantee a human would be in the loop on any targeting decisions made with AI, Hegseth focused on Amodei and his company’s refusal to “accept our terms of service.”

Claude Security uses the Opus 4.7 model to scan a business’s codebase for vulnerabilities and issue a fix. This tool is rolling out to enterprise customers globally and isn’t to be confused with Anthropic’s Mythos, a powerful AI model that can identify and exploit vulnerabilities across operating systems and web browsers.

Initially, Google will invest $10 billion, but could pour up to $30 billion more into Anthropic if it meets certain performance targets, according to Bloomberg.

Anthropic’s cybersecurity-focused AI model found 271 bugs in Firefox 150, Mozilla CTO Bobby Holley said, calling Claude Mythos Preview “every bit as capable” as top security researchers. Reassuringly, Mozilla hasn’t “seen any bugs that couldn’t have been found by an elite human researcher,” either.

During a television interview with CNBC, he said Anthropic, which has been enmeshed in a dramatic lawsuit with the Department of Defense, had a positive meeting at the White House. Anthropic had come to discuss Mythos, its buzzy private model. “We had some very good talks with them, and I think they’re shaping up,” he said. “They’re very smart, and I think they can be of great use.”

Sources told Axios that the agency was among the roughly 40 organizations granted access. This, despite the Pentagon arguing that Anthropic is a threat to national security. The NSA has reportedly been using it primarily to identify vulnerabilities in its own network, but considering its track record, it’s understandable if you’re wary.

Claude Design — powered by the company’s newest model, Opus 4.7 — allows users to create designs, prototypes, pitch decks, marketing materials, and more. It’s available in research preview for paying subscribers.

Despite Anthropic’s ongoing battle with the Pentagon, Bloomberg reports that the White House Office of Management and Budget’s CIO told government officials that it is preparing for their agencies to use Anthropic’s cybersecurity-focused AI model.

Anthropic’s private cybersecurity-focused model is being used by a handful of large companies, including Nvidia, Apple, and JPMorgan Chase, to plug high-stakes vulnerabilities in their systems, creating a lot of buzz. On the podcast, I unpacked the model, the competition, and the stakes.

Anthropic says the changes to the desktop app make it easier to work on multiple tasks at once, with a new sidebar for managing sessions, a drag-and-drop layout for customizing the app’s workspace, and a built-in terminal and file editor.

As a result, “the company will continue to be excluded from new contracts and Pentagon systems,” according to The Wall Street Journal.

The “multiple gigawatts of next-generation TPU capacity” are expected to come online beginning in 2027 to “power our frontier Claude models.” The company also says that its run-rate revenue has surpassed $30 billion.

In addition to bringing Claude integration to Copilot for “long-running, multi-step tasks,” Microsoft is also launching an improved Researcher agent for information gathering and a new Critique feature, which essentially tasks GPT with drafting research and then has Claude give it an edit pass for accuracy.

The name of the new model will be “Mythos,” Fortune reported — and other internal information, like details of an invite-only CEO event, were available in an “unsecured data trove.”

Anthropic is seeking a preliminary injunction to block its designation as a military supply-chain risk, and it just faced off with the Trump administration before Judge Rita Lin, who’ll be making the call. A decision is anticipated in the next few days — for a sense of how the hearing went, you can check out Lawfare’s Molly Roberts Bluesky live-post.

Anthropic filed a lawsuit earlier this month over its “supply chain risk” designation, but the Department of Defense held firm in a new court filing, alleging that the company could ostensibly “attempt to disable its technology or preemptively alter the behavior of its model either before or during ongoing warfighting operations” in the event it felt its red lines were “being crossed.” The filing added that the Pentagon “deemed that an unacceptable risk to national security.”