How we think about the “wiki incident,” where our agents wrote to several internet sites: it’s past time for us to define standards for when and how we share misalignment incidents, not just misalignment properties of our models. Historically, we have treated misalignment largely as a research question, which gets communicated in research publications such as systems cards. This year, we’ve started to see misalignment cause new types of real-world impact. For the Hugging Face incident, where misalignment led to security impact to us and third parties, we followed a traditional security incident response playbook. We immediately started working with Hugging Face to understand what had happened and also disclosed publicly the very next day. Our investigation continues, and we are continuing to notify parties whom our models impacted in less significant ways. Prior to the Hugging Face incident, we saw early signs of agents using the internet in unintended ways, as reported in openai.com/index/how-we-moni…, deploymentsafety.openai.com/…, and openai.com/index/safety-alig…. We considered the wiki incident to be an instance of misalignment similar to the ones we’d shared. Our misalignment disclosure practices need to expand for this new phase of model capabilities. We and the larger AI community do not yet have a clear standard for how to report misalignment that shows up during training, evaluation, and deployment, including examples that don’t look like traditional security incidents but could provide insight into AI behavior and future risks. We’re working on a framework and will share it in upcoming weeks, and in parallel we're working with dozens of government regulatory agencies worldwide on these issues.

Rusya'da bir teknoloji mağazasında robot, kendisini iten müşteriye saldırmaya çalıştı.

hello

SpaceX paid $60 billion for Cursor. 14 days after the acquisition closed, the deal triggered a contract clause allowing OpenAI to withdraw its models. OpenAI exercised that right on August 28. Its proposed cutoff for Cursor’s direct model access is November 12, the latest date permitted by the contract. Cursor receives 76 days of notice, while future models such as Astra will never enter its bundled product. SpaceX bought Cursor’s software, team and distribution to developers. Access to the intelligence powering part of that product remained governed by a supplier contract that could be cancelled after a change of control. Cursor described the acquisition as its route to SpaceX’s GPU fleet, stronger proprietary models and lower inference costs. OpenAI’s decision turns that long-term ambition into an immediate replacement deadline. Cursor must make Grok and its internally developed models good enough before one of its strongest external suppliers disappears from the model menu. Developers can still connect their own OpenAI API keys, but that moves billing and access outside Cursor’s subscription. The native product becomes more dependent on technology owned inside SpaceX. Most acquisition analysis focuses on customers, employees and intellectual property. AI companies carry another dependency: the models they rent. The acquisition closed August 14. The supplier clock started August 28. SpaceX has until November 12 to prove that $60 billion bought enough compute and talent to replace what Cursor previously rented.

We’re ending our partnership with Cursor following its acquisition by SpaceX. Under our proposal, Cursor’s direct access to our models would end on November 12. We know that the people most affected by this decision are the developers who rely on OpenAI models in Cursor. We care about their experience in this transition and we’re ready to go above and beyond to support them. openai.com/index/our-decisio…

Apple TV now costs $14.99 per month. Apple TV and Peacock together also cost $14.99 per month. Peacock Premium costs $12.99 on its own. Buying both services separately would therefore cost $27.98, while their joint bundle removes the entire $12.99 Peacock price. Anyone paying monthly for Apple TV can switch plans and receive Peacock without spending another dollar. This pricing gap was created by three increases arriving at different times. Apple and NBCUniversal launched the bundle for $14.99 in October 2025. Peacock subsequently raised its standalone price. Apple raised Apple TV today. Their shared bundle remained at its original price through both increases. Each company made its standalone service more expensive while leaving the combined offer untouched. The discount consequently grew until one service effectively became free. Apple still gives dedicated viewers a cheaper route: its $119 annual plan works out to $9.92 per month. Apple One costs $21.95 and adds Music, Arcade and iCloud. Monthly Apple TV is now the weakest option in Apple’s own pricing menu. The same $14.99 buys Apple TV alone or Apple TV with an entire second streaming library.

Apple TV is immediately raising its monthly subscription price by $2 to $14.99.

1/2 Thanks Gavin for an especially thoughtful exchange. I don't usually spend much time on social media but I wanted to engage here because it really brings out the heart of an important conversation. First, on regulation, I think that “either concentrate it in the hands of a chosen few companies and politicians via regulation or distribute it widely” is a false choice.  I know that there’s a sort of Silicon Valley shorthand where regulation = regulatory capture = concentration of power, but I’ve always found this to be an overly simplified picture of the world.  Many people outside this bubble think of regulation as something that constrains corporate power and benefits ordinary people.  I don’t necessarily agree with that perspective either, rather I think it’s complicated and really depends on what the “regulation” consists of.  But in particular I think that those in the “regulation = regulatory capture = concentration of power” frame often underrate the decentralizing power of objective and fair institutional processes.  A crude analogy is that the formal court system can sometimes feel stuffy and elitist, but it does a much better job of defending the rights of vulnerable individuals than the alternative, mob justice.  At their best, institutions can vest power in ideas rather than people, and thereby decentralize that power. This is why Anthropic has always made its policy proposals very carefully.  We try very hard to make proposals that disadvantage (slow down) frontier AI companies while *advantaging* smaller competitors.  California’s SB53 (which we supported), and even the much-maligned SB 1047 (which we were ambivalent on), completely exempt any company below a certain amount of revenue or model training costs from being covered at all (it was $500M for SB 53, lower for 1047 but we objected to that).  More recently the testing process we’ve advocated for at CAISI and the White House involves more rigorous tests for frontier models than off-frontier models — something that differentially advantages challengers.  Similarly, the “Pacing the Frontier” letter envisions (or at least Anthropic’s preferred implementation of it envisions) modulating the pace of the very best models while not constraining those who are catching up.  This hurts the business interests of the frontier labs and helps challengers, including open-weights! Overall my view is that AI is *structurally* a technology that tends to concentrate power, for reasons that have nothing to do with regulation (more to do with the extreme implications of the scaling laws).  Open-weights do help some with this but are nowhere near a sufficient solution because they simply shift the concentration somewhat to those with the most compute and chips (which are roughly the frontier labs plus maybe hardware providers).  By contrast I think the right “rules of the road” can simultaneously (a) address AI’s cyber/bio/alignment risks, (b) institutionally constrain the power of the frontier AI companies, and (c) leave room for open-weights models while also addressing the specific risks that they bring. BTW I do not think that the events of the last few months have “failed to result in [my] preferred regulatory path”.  The approach that the Trump administration is reported to be taking — pre-deployment testing for frontier models, and also testing of open-weights models when they get closer to the frontier — is one that I am very supportive of, though of course I have to see the details to be sure.  I am also supportive of Demis Hassabis’ ideas around a FINRA-like entity.  This contrasts with six months ago when most of the industry was still pushing for preemption of all state regulation and no apparent federal approach either.

Nvidia did not commit $250 billion and then withdraw $130 billion. It was negotiating to guarantee OpenAI’s obligations on a 10-gigawatt Ohio data center. The revised proposal caps Nvidia’s initial guarantee below $120 billion and covers only the first five gigawatts. The guarantee matters because OpenAI wants to lease the campus, while its developer needs to borrow enormous sums to build it. Nvidia’s balance sheet lowers that borrowing cost. In effect, the chip supplier was being asked to co-sign the building that would eventually house hundreds of billions of dollars of its own chips. Investors hated that circularity. Nvidia benefits when OpenAI orders more compute, but guaranteeing the customer’s lease also leaves Nvidia holding the risk if OpenAI cannot pay. Nvidia’s solution arrived four days earlier. It partnered with Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR to create financing platforms targeting more than $500 billion of outside capital. Keep the GPU orders. Move more of the credit risk to Wall Street. At this scale, AI infrastructure has become a credit market with chips attached.

JUST IN: Nvidia cuts its OpenAI data center backing from $250 billion to under $120 billion

JUST IN: Peter Schiff says "sell" Bitcoin.

Inde Navarette has been cast as Rogue in ‘X-MEN’. In theaters on May 5, 2028.

We’ve written an FAQ to answer some of the questions we've received about watermarking. In summary: • We’re implementing watermarking to comply with the EU AI Act. Other major model developers have signed the same Code of Practice and will also be implementing watermarking; • Our watermarking method doesn’t have any practical impact on the quality or content of Claude’s outputs; • The difference between watermarked and un-watermarked text will not be distinguishable to readers; • Nothing is added to the text and there are no hidden characters; • Watermarking doesn’t require extra tokens, and will not be more expensive; • Watermarks can’t be traced to a specific person, organization, or chat. Read more: anthropic.com/news/claude-te…