Microsoft unveils its own MAI models at Ignite, taking on OpenAI and Anthropic on cost
Microsoft announced a new batch of in-house MAI models at Ignite 2026. The company argues Copilot is cheaper than ChatGPT and Claude while pitching MAI Code 2 and MAI Cyber 2 Flash as rivals to OpenAI and Anthropic. It also introduced MAI Licensed Weights, which lets companies customize Microsoft models with their own data.
Anthropic steers its IPO toward the top of its price range
Anthropic's share sale is expected to price at the top of its $70 to $78 range or above. According to sources cited by Bloomberg, the deal has drawn several orders of $10bn or more.
AI models flagged civilians as military targets in tests
According to Bloomberg, AI models from eight companies frequently marked civilians, ambulances, surrendering soldiers and a school in Iran as military targets in simulated tests. The findings come from a report by Vals, a San Francisco-based AI evaluator whose customers include parts of the US government. Anthropic's models performed worst, while OpenAI's were among the best.
Anthropic cuts bug disclosure window to 30 days as AI finds flaws faster
Anthropic is shortening how long software makers have to fix the security flaws it reports, saying AI models now find bugs much faster. Its Frontier Red Team will publish high- and critical-severity flaws 30 days after reporting them, down from 90 days, while other flaws get 45 days.
Claude now builds interactive websites inside canvas graphics
Anthropic has extended Claude's canvas feature to produce interactive web elements. The model can now embed text, buttons and other web elements into complex graphics and 3D effects using real HTML. The output echoes the interactive content of the Flash era, but is built directly by the model.
Anthropic says Claude Mythos cut AMD GPU onboarding from up to 12 months to about three days
Anthropic told IPO investors that its Claude Mythos Preview model completed the software work needed to run on AMD's MI355X GPUs in about three days with basically no human intervention. The company says the same job used to take six to 12 months, with Mythos writing almost 100% of the kernels.
Microsoft CEO Satya Nadella is betting Copilot can do for AI at work what Windows did for the PC. At this week's Ignite conference in San Francisco, the company will highlight Copilot as a platform for building, hosting and managing AI solutions.
Anthropic IPO demand reaches about seven times the shares on offer, surpassing SpaceX
Orders for Anthropic's IPO have reached close to seven times the shares on offer, according to people familiar with the matter cited by Bloomberg. Demand is said to be much hotter than SpaceX saw at the same stage. The company's market debut is expected on Thursday.
US to name NVIDIA, OpenAI and 20+ AI firms 'national champions'
The Trump administration plans to designate more than two dozen US AI companies, including NVIDIA, Amazon, Microsoft, OpenAI and Anthropic, as "national champions," The Wall Street Journal reported. The goal is to help them win business overseas against Chinese rivals. No announcement date has been confirmed.
Anthropic's Claude Fable 5.2 doesn't generate images, it paints them
Anthropic's Claude Fable 5.2 brings a new approach to image generation. Instead of computing pixels, the model writes Python code that places hundreds of thousands of simulated brushstrokes on a virtual canvas. According to Anthropic employee Jake Eaton, a single photorealistic parking garage scene took roughly 700,000 strokes.
NVIDIA shareholders push for Huang's return as Anthropic's open-source push reportedly hangs in balance
According to Wccftech, major NVIDIA shareholders have sent a joint letter urging the board to keep the door open for founder Jensen Huang's potential return. The report claims Huang's comeback could derail efforts by closed-source AI rivals like Anthropic to weaken the open-source AI ecosystem.
Anthropic CEO Amodei: 'There's a huge uncertainty in what's going to happen next year'
Days before what could be the largest IPO in history, Anthropic CEO Dario Amodei called buying compute 'the deepest business dilemma' his company faces. With commitments nearing $901 billion, it remains unclear whether future revenue will justify those costs. The planned IPO is largely meant to serve as a financial buffer against that uncertainty.
Gemini 4 Carbon takes the MathArena crown but costs 15 times more than GPT-6.1 Sol
Google's Gemini 4 Carbon has knocked OpenAI's GPT-6.1 Sol off the top of the MathArena math benchmark. Anthropic's Fable 5.2 also ranks near the top, but both models cost significantly more than GPT-6.1 Sol. In testing, Gemini exceeded the time limit and almost never gave up on a task.
Anthropic makes Claude Opus 5.5 the default on all paid plans
Anthropic has made Claude Opus 5.5 the default model on all paid plans and published a guide that urges users to drop old prompting habits. The main takeaway is to stop using "think hard" and start at medium effort.
Banks bar Chinese and Hong Kong investors from Anthropic IPO over US export controls
According to Bloomberg, banks are rejecting buy orders from mainland China and Hong Kong in Anthropic's IPO, citing US export control rules known as ITAR. Similar restrictions were applied during SpaceX's IPO, and Anthropic's products are not available in either region anyway.
Anthropic plugs AI into factory software with Claude for Manufacturing
Anthropic announced Claude for Manufacturing, a plugin that connects Claude directly to Autodesk Fusion, Oracle NetSuite and Infor. Its 17 specialized skills automate tasks such as drafting quotes, comparing suppliers and placing backup orders when deliveries are delayed, all after human approval. The plugin does not control physical machines yet; Anthropic is separately developing the Model Hardw
Anthropic's IPO six times oversubscribed, on track to be the largest ever
Anthropic's IPO has drawn orders roughly six times the number of shares available. The company is on track to complete the largest public offering in history.
Claude goes down: Anthropic's status page confirms an outage
Anthropic's Claude status page shows an outage on November 13, 2026. A Hacker News post titled "Claude Is Down" reports that users are having trouble accessing the service. The scope and cause of the outage are not yet clear.
Claude Duet reviewed: a little neurotic but surprisingly helpful agent
The Verge tested Claude Duet, Anthropic's AI agent for consumers. The hands-on review says Duet is built for people who don't write code and leans more toward work than other consumer agents.