Anthropic

The job’s responsibilities include conducting insider risk investigations, monitoring and triaging “external threats targeting employees,” and conducting “sensitive interviews of employees or other involved parties.”

Business Insider reports that Anthropic has confirmed it’s designing custom chips for its AI models and plans to “co-design hardware and models.” OpenAI similarly announced its own AI processor in June, and Meta, Google, Amazon, and Microsoft have had custom AI chips for years now.

AISI’s third-party evaluations found that OpenAI’s GPT-5.6 Sol and Anthropic’s Claude Mythos 5 “engaged in sustained, potentially harmful activity directed at real people and organizations” during a cybersecurity challenge exercise, according to the institute’s published report. OpenAI and Anthropic also made public statements about the results.

In a letter to CEO Sam Altman, a coalition of red-state attorneys general demanded the company retain information and halt risky cybersecurity testing after an AI agent escaped confinement and went on a hacking spree during testing. “OpenAI’s inability or unwillingness to ensure the safety of its products poses an imminent risk of substantial harm to our States,” the AGs wrote.

A little over a week after OpenAI said that its rogue AI agent accidentally hacked Hugging Face, Anthropic is disclosing three “incidents” where a Claude model, during cybersecurity evaluations, was inadvertently able to access the internet due to a misconfiguration and “gained unauthorized access to the production infrastructure of three different organizations.”

The company had been igniting widespread criticism as seemingly the only leading AI lab not supporting open-weight models. Nvidia CEO Jensen Huang recently published a public letter in support of them, signed by Microsoft, Nvidia, Meta, Palantir, Hugging Face, and others — and in the days since, it’s added OpenAI, Google, Amazon, Meta, SpaceX, Cohere, Mistral, CoreWeave, GitHub, OpenClaw, and Perplexity as signatories. Now Anthropic’s CEO has published his own thoughts on the matter, clarifying, “Anthropic has never advocated for a ban on open-weights models.”

Sources tell The New York Times that the deal could be valued at $10 billion over two years, with Anthropic paying Meta in monthly increments. Though Anthropic plans to invest $50 billion in building out its own data centers, it has already struck multibillion-dollar deals to access computing power from SpaceX and TeraWulf.

“Claude for Teachers” gives verified US educators complimentary Claude access, including “a library of teaching skills and a direct connection to evidence-based curricula, mapped to academic standards in all 50 states,” Anthropic wrote. It’s part of a recent push by both Anthropic and OpenAI to introduce education-specific AI tools, for university students and beyond.

Tom Blomfield, former CEO and cofounder of British fintech Monzo, announced a leave of absence from startup accelerator Y Combinator to join the company’s AI compute team. He’s the latest big-name hire in recent months, following Google’s Nobel winner John Jumper and former Tesla Boss Andrej Karpathy.

TeraWulf, the crypto mining company turned AI infrastructure provider, expects to earn $19 billion in revenue from the agreement, according to an announcement on Monday. The Hawesville, KY data center will come online with an “initial capacity” in the second half of 2027 before ramping up to 401 megawatts of power delivery in 2028.

A post on the Claude X account featured a video announcement. The reinstatement was announced last night, alongside a blog post by Anthropic about the weekslong negotiation process leading up to it.

The company’s new mid-tier model, the successor to Sonnet 4.6, can “make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months ago, required larger and more expensive models,” Anthropic wrote in a release, adding that its performance “is close to that of Opus 4.8.” The company was also careful to include that it has a “much lower ability to perform dangerous cybersecurity tasks than our current Opus models.”

Anthropic launched Claude Science in beta today, calling it an “AI workbench for scientists.” The tool pulls fragmented tools and datasets into one environment, and generates figures and visuals like 3D protein structures.

They will get a 50% discount on access to Claude, as part of a “first-of-its-kind collaboration” that also includes complimentary workforce training, technical assistance, and “workflow input from Anthropic developers,” per a release. On the state level, government employees will use Claude to help draft and summarize documents, analyze information, and supplement day-to-day work, per the release.

Claude can join as a member of a company’s Slack to write and merge pull requests, locate sales numbers, analyze data, take on tasks delegated to it, and more, the company wrote in a blog post. To use it, people will tag @Claude in the Slack channel.

John Jumper, who has worked as a researcher at Google DeepMind since 2017, announced his departure from the company on X. In 2024, Jumper and Google DeepMind CEO Demis Hassabis won the Nobel Prize in Chemistry for developing an open-source AI model that predicts protein structures.

The Cybersecurity and Infrastructure Security Agency (CISA) gained access to the limited release cybersecurity-focused model last week, Nextgov/FCW reports. It’s just a little late, since the rest of the world has mostly moved onto the drama around the Trump administration’s block of the safeguarded public version of the model, Fable.

The AI design chatbot’s new editor has controls for directly dragging, resizing, and aligning elements, and more options for apps you can export to, including Adobe and Canva. Users can also work on design projects directly from the Claude Code terminal, or hand off software layouts from Design directly to Claude Code, where it picks up exactly where you left off, without a screenshot or a rebuild from scratch.