OpenAI says Astra crossed its critical cyber threshold

OpenAI plans to release Astra soon, but says it will restrict the model's most advanced cyber capabilities to selected partners at launch.

Read the full story →Listen to the edition

OpenAI ↗WIRED ↗TechCrunch ↗

Listen to this edition

Prefer audio? The briefing has chapters and a full transcript.

Follow on Apple Podcasts Follow on Spotify Transcript All episodes →

The briefing

The rest of the morning

4 more stories

Models, coding, pricing, privacy, safeguards, and controlled accessAnthropic paired a general model with a restricted version for higher-risk cyber and biology work.

Anthropic released Fable and Mythos 5.1

Anthropic released Claude Fable 5.1 for general use and Claude Mythos 5.1 through trusted-access programs. The company says both versions use the same underlying model with different safeguard levels. Anthropic estimates that Fable 5.1 will cost 25 percent less than Fable 5 on typical token-billed work. It estimates savings up to about 45 percent on highly agentic work because cache-read prices are lower. Anthropic also says its new cyber safeguards produce 60 percent fewer false positives. Enterprise Frontier Safeguards are scheduled to roll out in phases later this fall. Anthropic says they keep monitored data in customer-controlled infrastructure.

The price, performance, and safeguard figures come from Anthropic, not independent tests. TechCrunch reported that Anthropic's system card calls Mythos 5.1 a slight regression from Opus 5 on overall misaligned behavior. TechCrunch also reported that the card shows improvement over Mythos 5 and Sonnet 5. Fable can help discover software flaws, but Anthropic does not mean it to develop exploits. Mythos access rules and the planned enterprise privacy system still need outside testing.

Anthropic ↗TechCrunch ↗

Read article →

Politics, international rules, industry influence, competition, and public oversightThe United States asked major economies to use existing rules for most AI issues instead of building new oversight bodies.

The United States pressed G20 members to avoid new AI regulators

At a G20 technology meeting in Chapel Hill, North Carolina, White House technology adviser Michael Kratsios urged members to adopt the Carolina Principles. Reuters reported that the principles ask governments to save new regulation for novel issues. They also call for investment in foundational research and more commercial opportunities for new technology. A White House official said the United States would also press members not to create new organizations solely to oversee AI. The position follows the administration's March call for one national framework instead of conflicting state rules.

The principles are a policy proposal, not binding law. No complete public text or member-by-member record shows what governments accepted. The approach fits the interests of large American AI companies, which benefit from fewer new restrictions. Existing sector rules can cover some harms. The administration has not shown that those rules can inspect frontier-model access, autonomous cyber behavior, or cross-border incidents at the required speed.

Reuters ↗The White House ↗

Read article →

Products, video analysis, developer costs, benchmarks, and model efficiencyGemini can now choose which parts of a video to inspect instead of processing every segment at one fixed frame rate.

Google launched agentic video analysis in Gemini

Google launched agentic video understanding for Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite. The feature lets the model search, resample, and inspect selected video segments across frames, audio, and transcripts. Google offers it for uploaded and YouTube videos through the Gemini API in Google AI Studio and the Gemini Enterprise Agent Platform. Google says its tests cut token use by up to 88 percent. It also says cost fell by up to 66 percent, while accuracy improved by up to 7 percent.

The efficiency and accuracy figures are company claims. Google did not publish an independent audit, complete task-level results, or one guaranteed improvement across all videos. The largest gains are upper bounds across selected benchmarks. Dynamic inspection can also miss the right segment if the search step fails.

Google ↗

Read article →

Companies, cloud contracts, chips, data centers, power, and financingAnthropic's reported contract ties one model company's growth to a large cloud and data-center chain backed by Nvidia.

Anthropic reportedly signed a $35 billion Lambda cloud contract

Reuters and The Wall Street Journal reported that Anthropic signed a $35 billion cloud-computing contract with Nvidia-backed Lambda. The reported capacity will come from a Hut 8 data center under development in Nueces County, Texas. Reuters reported about 350 megawatts of capacity. The Wall Street Journal reported that Nvidia holds the lease on the site, while Lambda will provide Anthropic with the compute.

Unnamed sources described the contract value and structure because the terms are private. Anthropic, Lambda, Nvidia, and Hut 8 had not publicly confirmed the reported agreement when the stories were published. A signed compute contract does not prove the site is ready. It does not show finished construction, every power milestone, installed chips, or Anthropic traffic.

Reuters ↗The Wall Street Journal ↗

Read article →

The watchlist

Signals that could change the read

WatchlistAstra's release date, independent exploit tests, and Daybreak Blue access records.Tracking

WatchlistActual Fable 5.1 bills, safeguard false positives, and Mythos enrollment rules.Tracking

WatchlistA public Carolina Principles text and member-by-member commitments.Tracking

WatchlistCompany confirmation and delivered capacity for the reported Anthropic-Lambda contract.Tracking

Across the desks The leading AI companies are releasing stronger systems. They are also tightening access, cutting inference costs, and signing much larger compute contracts. Their own tests are moving faster than independent checks and public rules.

4 sourced signals frame today’s briefing.

SecurityOpenAI says Astra reached its critical cyber thresholdThe company has not released Astra or independent evidence for its strongest capability claims.

ModelsAnthropic released Fable and Mythos 5.1The price, performance, and safeguard improvements come from Anthropic's own tests.

PolicyThe United States urged G20 members to limit new AI regulationThe Carolina Principles are nonbinding, and no complete public text shows what each country accepted.

InfrastructureAnthropic reportedly signed a $35 billion Lambda cloud contractThe parties have not publicly confirmed the private terms reported by Reuters and The Wall Street Journal.

Editorial direction, not a financial index. Each signal is tied to this edition’s reporting.

Edition validated · 5 stories · 10 unique sources

About HotTea & our sources →