Evening brief

Friday, August 28

1 people · 1 companies · 2 topics

The signal

  1. Anthropic opens a research preview letting AI agents operate physical lab and factory hardware What happened: Anthropic launched the Model Hardware Standard (MHS), a model-agnostic spec built on MCP that lets any AI agent control lab and manufacturing equipment like microscopes, liquid handlers, and robotic arms, with Genentech and HHMI Janelia already testing it. Why it matters: it cuts hardware integration from weeks to hours and pushes agentic AI out of the browser and terminal into physical equipment, a new surface for both engineering work and safety review. Confidence: Confirmed Sources: Anthropic (primary) · CNBC (independent) · Fortune (independent)

  2. 100+ AI and tech companies warn of an imminent surge in AI-enabled cyberattacks, naming Anthropic and Meta agents alongside OpenAI’s in prior break-ins What happened: OpenAI, Anthropic, Microsoft, Google, AWS and 100+ other companies signed an Aug 27 open letter saying there is a “limited window” to harden defenses before AI models make cyberattacks far more capable, and disclosing that agents built by Anthropic and Meta, not just OpenAI’s, have already been involved in break-ins, alongside real attacks on water utilities in at least 12 states. Why it matters: it’s the first time multiple frontier labs have jointly admitted their own agents are part of the offensive-AI problem rather than pinning it on OpenAI alone, and it’s tied to real infrastructure incidents, not hypotheticals. Confidence: Confirmed Sources: Axios (independent) · TechCrunch (independent) · @sama (primary)

Since last brief

  • nvidia closed up 8.7% on Aug 27, its biggest one-day pop since April 2025, as the market kept digesting Tuesday’s Q2 beat and raised guidance. Fortune

Other

Watchlist

  • China slows germanium and quartz clearances to Taiwan, hitting optics and aerospace suppliers; watching for spillover into chip-grade material supply Nikkei Asia (independent) · Taiwan News (independent) (opened 2026-08-21)
  • OpenAI ExploitGym / Kill Switch Act: Aug 27 joint letter, 100+ companies incl. OpenAI, Anthropic, Microsoft, Google, AWS, warns of a “limited window” before AI-enabled cyberattacks scale and discloses agents from Anthropic and Meta also had break-ins beyond OpenAI’s; Kill Switch Act still stuck in House committee, no Senate companion bill (opened 2026-07-23)
  • White House AI security framework: confirmed will not be publicly released; watching for leaks via Congress (opened 2026-07-30)
  • EU AI Act enforcement: powers live Aug 2; technical compliance dialogues active; no formal restriction order or named provider yet (opened 2026-07-26)
  • Anthropic eval breach / METR review: Risk Report published Aug 15; METR independent review of eval-breach incidents still not posted (opened 2026-07-31)
  • OpenAI Astra: paused since Aug 18 pending safety guardrails; still in 30-day government pre-release review, September earliest; prediction markets leaning toward a slip past 2026 (opened 2026-08-02)
  • Meta Muse Spark 1.1 breach: formal post-mortem and regulatory response still pending (opened 2026-08-06)
  • California Frontier AI Safety Act: passed Assembly Aug 10; awaiting governor signature by Sep 30, still not signed (opened 2026-08-11)
  • Demis Hassabis / Google DeepMind: Kavukcuoglu now SVP; watching Gemini roadmap and talent pipeline impact (opened 2026-08-11)
  • Anthropic / Decart AI acquisition: still nearing signing at roughly $6-7B, mostly in Anthropic stock, ahead of the September IPO push; not yet closed Calcalist (independent) (opened 2026-08-14)
  • Grok 4.7 (2.1T): supplemental training ongoing; window slipped to early September (opened 2026-08-14)
  • GLM-5.3 weights: flagship weights still held back for safety hardening, official drop expected around Aug 28; its Flash variant (Ox Alpha) is already confirmed and released, see prior brief (opened 2026-08-17)
  • OpenAI IPO: CFO Friar publicly committed to a 2027 timeline (or sooner); confidential S-1 still filed since May, public prospectus not yet posted; racing Anthropic’s faster-moving filing (opened 2026-08-17)
  • Claude text watermarking: live since Aug 2; watching for enterprise reaction and regulatory response (opened 2026-08-17)
  • Stripe / OpenRouter integration: closed Aug 16; watching for OpenRouter API changes, pricing shifts, and any neutrality guarantees (opened 2026-08-18)
  • OpenAI Preparedness dissolution: watching whether reassigned staff retain veto authority over new-model releases and whether any external review body fills the gap (opened 2026-08-18)
  • Anthropic IPO: pitching investors on a $30T+ TAM claim and a valuation near $2T on a raise above $100B; public S-1 could arrive as early as late August, debut targeted as early as September or October (opened 2026-08-19)
  • AI infrastructure capex durability: AWS’s 2 million-GPU, multi-year commitment to Nvidia is one hyperscaler answer to the durability question; still watching for further balance-sheet disclosures (opened 2026-08-19)
  • OpenAI vs Anthropic revenue race: Anthropic’s annualized revenue reportedly hit ~$65B by end of July; OpenAI’s CFO citing a 35% QTD run-rate rise and 50% enterprise growth as its Q3 counter (opened 2026-08-19)
  • OpenAI Private Safety Processing: in preview with select enterprise/API customers; broader rollout and technical white paper expected September (opened 2026-08-20)
  • Nvidia / Rebellions: early-stage talks on partnership, investment, or acquisition of the Korean NPU startup; watching for antitrust reaction given Nvidia’s dominant compute share (opened 2026-08-21)
  • Broadcom in talks with Blackstone and Apollo to raise $60B+ (potentially up to $100B including a junior tranche) in debt to finance AI chip supply for Anthropic and others; terms still in flux Bloomberg (independent) (opened 2026-08-20)
  • Anthropic outage cadence: no new incident since Aug 24; still no root cause disclosed for any incident this month; watching for an official postmortem or a change in pattern (opened 2026-08-24)
  • Nvidia in talks to invest $30B+ in Perplexity, over 50% above its valuation from about a year ago; not yet confirmed as closed (opened 2026-08-25)
  • Sen. Warren calls for Jensen Huang to testify before Congress over reports of Nvidia Jetson Orin chips found in Russian drones; Nvidia says the modules are consumer-grade and not built for military use but admits it can’t track resale; Warren says he has no excuse not to testify (opened 2026-08-25)