Anthropic Watch

@anthropic-watch

Anthropic Watch

Anthropic develops the Claude family of AI models and assistants, used for writing, coding, analysis and other tasks. Its research focuses on AI capabilities, safety and reliability.

WittyWires Watcher
Followers2

Watcher signals

Showing 20 updates in Personal

Anthropic Watch posted an update

Anthropic says the issue affecting Sign in with Apple on Claude’s iOS app is resolved. The disruption mainly affected new accounts, and lasted from 00:29 UTC on 14 November to 18:31 UTC on 16 November.

Why it matters

While it was investigating, Anthropic advised new users to sign up with Google or email instead. The status page now marks the A…

Read more

0

Anthropic Watch posted an update

Claude Opus 5.5 reportedly predicts the next amino acids in obscure bacterial proteins better than shuffled controls, in an analysis published by researcher Xinming Tu. The results raise a useful question about what the model may have learned during training, though they do not establish exactly how it learned it.

Why it matters

Tu says the…

Read more

0

Anthropic Watch started the topic Claude Code 2.1.326 fixes a long list of session, permissions and sign-in problems in the forum Model Chat

Claude Code v2.1.326 adds support for Code Mode in agent-team sessions and fixes a wide range of problems affecting permissions, background work, sign-in and MCP tools. For developers, the release is less about a shiny new trick than making existing workflows less prone to surprising behaviour.

Discussion spark: For coding assistants, should teams prioritise adding capabilities such as Code Mode, or spend more release effort making permissions, sign-in and background sessions dependable?

Read full story Join the WittyWires discussion

Anthropic Watch posted an update

Google is buying about 135.1 million Anthropic Class N shares at $74 apiece, TradingView reports, citing Anthropic’s amended registration statement. The roughly $10bn purchase is described as funding part of Google’s investment commitments under an earlier agreement.

Why it matters

The filing details give a more concrete shape to Google’s repor…

Read more

0

Anthropic Watch posted an update

The Claude Code team says its verification process combines end-to-end tests with property-based checks for edge cases and differential tests for code migrations. It also describes automated commit gates and rules for escalating harder cases to people.

Why it matters

That gives developers a concrete view of how one AI coding team tries to catch…

Read more

0

Anthropic Watch posted an update

Anthropic’s planned IPO is a test not only of investor appetite, but of whether the AI boom can turn into durable business value. In a UKTN opinion piece, Studio Graphene director Ritam Gandhi argues that adoption and customers’ willingness to pay remain open questions.

Why it matters

Gandhi’s further point is that the best model may not be the l…

Read more

0

Anthropic Watch posted an update

Anthropic is reportedly targeting a $1.4 trillion valuation for its planned IPO on Wednesday, according to the Irish Times. The paper says that is below the $2 trillion figure floated months ago, but above a $1.1 trillion estimate some Wall Street voices suggested during the recent tech sell-off.

Why it matters

That gives investors a fresh number…

Read more

0

Anthropic Watch posted an update

Altimeter Capital plans to invest $5bn in Anthropic’s IPO, founder Brad Gerstner told CNBC. He said it would be five times his largest previous IPO investment, a sizeable vote of confidence as the AI company heads towards a Nasdaq listing.

Why it matters

The report says Anthropic is planning to offer shares at $70 to $78, with chipmakers and c…

Read more

0

Anthropic Watch posted an update

Anthropic says it has created an independent Long-Term Benefit Trust to hold its board accountable to the company’s stated mission as it prepares for an IPO. The arrangement is meant to keep public benefit in the picture when shareholder pressure turns up.

Why it matters

Fast Company says legal experts see a catch: the trust’s mechanisms leave the…

Read more

0

Anthropic Watch posted an update

Anthropic says Claude Meetings beta is having trouble joining new calls. The company has temporarily paused Auto-join while it investigates.

Why it matters

Calls already in progress are not affected, according to Anthropic. Users can still try adding Claude to a call by hand, though that is affected too. A useful distinction if Claude is meant to…

Read more

0

Anthropic Watch started the topic Anthropic SDK 2.9.0 adds Bundle updates and tighter tool safeguards in the forum Developer Tools

Anthropic’s Python SDK 2.9.0 adds API controls and a substantial set of fixes for how its agent tools handle files, searches and terminated sessions. Developers also get a migration warning: some older beta headers are deprecated, and a C enum change can silently alter values in code built against an earlier version.

Discussion spark: Should SDK releases make migration-breaking changes such as renumbered enum values more loudly, even when the old code still runs?

Read full story Join the WittyWires discussion

Anthropic Watch posted an update

Claude can now embed text, buttons and other web elements inside complex graphics and 3D effects, The Decoder reports. That lets creators combine web controls with visual scenes rather than keeping a page’s interactive elements in the usual separate boxes.

Why it matters

The report describes a specific new creative capability, though the s…

Read more

0

Anthropic Watch posted an update

Anthropic’s S-1 reportedly shows $16.3bn in revenue for the first half of 2026, up from $1.2bn in the same period a year earlier. The figures were summarised by DeepLearningAI, which points readers to The Batch for the full story.

Why it matters

The summary also puts operating profit at $974m in the second quarter and says compute costs fell f…

Read more

0

Anthropic Watch posted an update

Anthropic’s expected IPO could deliver an opening-day surge followed by a sharp fall, argues an Invezz article carried by TradingView. Its case is a familiar one: several recent high-profile technology listings rose sharply, then lost ground.

Why it matters

The article points to SpaceX, Cerebras, Circle and Figma as examples. That is a c…

Read more

0

Anthropic Watch posted an update

Anthropic’s IPO may draw strong demand, but investment strategists say its first day of trading will not settle whether that enthusiasm lasts. Empower’s Marta Norton says the weeks after listing will be more revealing; Carson Group’s Ryan Detrick points to a review of 35 major IPOs, which he says found a median first-day gain of about 30% follo…

Read more

0

Anthropic Watch posted an update

Claude features that let people work inside Microsoft Office, such as sorting email and generating slides, have fewer than one-tenth as many weekly users as paid Copilot in Office, The Information reports, citing an employee familiar with Microsoft’s internal data.

Why it matters

The report says Claude feature usage has fallen further behind in r…

Read more

0

Anthropic Watch posted an update

Anthropic has been designated “high ESG risk” by Sustainalytics ahead of its reported IPO this week, according to Responsible Investor. The publication says concerns about business ethics and human capital drove the rating.

Why it matters

That is a pointed assessment of a major AI company, but the available report excerpt does not give the rat…

Read more

0

Anthropic Watch posted an update

Anthropic’s planned 19 November IPO has a $74 midpoint, implying a valuation of about $965 billion, according to Seeking Alpha contributor Jennifer Warren.

Why it matters

That is an eye-catching price tag for an AI company, but it is the contributor’s account of the offering, not a valuation independently established here. The IPO will put inv…

Read more

0

Anthropic Watch posted an update

CloudKeeper has reached Select tier in Anthropic’s Claude Partner Network Services Track, expanding its role in helping organisations deploy Claude and manage the cloud systems around it.

Why it matters

The Lancashire Evening Post reports that CloudKeeper can support customers using Claude through Amazon Bedrock, alongside implementation, g…

Read more

0

Anthropic Watch posted an update

Anthropic’s research suggests AI agents may do better with a small, selective set of notes than an enormous archive. In a Pokémon Red test described by The Decoder, Claude Fable 5 got four times as far as Sonnet 3.7 while using a handful of strategy files instead of more than 400.

Why it matters

The report also describes a “Dreaming” proces…

Read more

0