Sarmadi AI Digest July 2, 2026 Updated 7:00 AM CT Today Archive Topics Saved Subscribe RSS

OpenAI floats a 5% stake for Trump; Cloudflare forces AI-crawl payments; Google admits +37% electricity for AI

OpenAI reportedly floated giving the US government a 5% ownership stake as a way to ease federal friction — the most explicit political-economy proposal any frontier lab has publicly aired. Wired documents the new security measure Anthropic added to regain access, and Ars confirms the Mythos/Fable release is now global. Cloudflare is giving AI companies until September 15 to separate crawlers from paid indexing — the AI-publisher payment moment finally arrives. Google's environmental report shows its AI buildout drove a 37% increase in electricity use in 2025. Meta is building a cloud business to sell excess AI capacity, and separately started charging a subscription for smart-glasses features. Venice AI became a profitable unicorn with $70M ARR on a privacy-first model. Ashton Kutcher left Sound Ventures to start a new AI-focused VC with Morgan Beller.

6 papers 15 news 7 sources ← Latest

News

16 items

OpenAI floats a 5% federal stake; Anthropic gets global release

OpenAI reportedly floated a 5% government ownership stake to smooth the federal-AI relationship — the most explicit quid-pro-quo any frontier lab has publicly aired. Wired documents the new security measure Anthropic added to get back into the White House's good graces. Ars confirms Mythos/Fable now cleared for global release.

News The Verge AI

OpenAI floats giving Trump administration 5 percent cut of AI boom

The Verge: OpenAI has floated giving the US government a 5% ownership stake as a way to ease AI restrictions.

proposed stake 5%
Why it matters
  • First public frontier-lab proposal to hand real equity to the US government — reframes the entire pre-vetting debate.
  • Sets the counter-offer benchmark against Anthropic's new security-measure concessions.
  • Materially affects the political-economy argument the Sanders and Bores stories made real earlier this cycle.
  • Directly relevant to OpenAI's disclosure profile heading into any IPO conversation.
policyregulationmarket
Source →
News Wired AI

Anthropic Added a New Security Measure to Get Back Into the Trump Administration's Good Graces

Wired: Anthropic added a specific new security measure as a condition for the export-control lift on Fable 5 and Mythos 5.

Why it matters
  • Provides the concrete quid-pro-quo behind yesterday's export-control reversal.
  • Sets the technical-concession template other labs will be pressed to match.
  • Explains the speed of the four-day reversal — the political cost had a specific technical price.
policysafetyregulation
Source →

Cloudflare forces AI-crawl payments; Google's +37% electricity year; Meta pivots to cloud

Cloudflare gave AI companies until September 15 to pay for or drop crawls of publishers' content — one of the largest ecosystem-level forcing functions of the year. Google's environmental report shows the AI buildout drove a 37% increase in its 2025 electricity use. Meta is building a cloud business to sell excess AI capacity, following the SpaceX-style path.

News TechCrunch AI

Cloudflare's new policy pushes AI companies to pay for publishers' content

TC: Cloudflare will require AI companies to separate crawlers used for search vs training by September 15 — publisher payment moment.

deadline Sept 15, 2026
Why it matters
  • First infrastructure-layer forcing function on AI-crawl compensation with a real deadline.
  • Aligns with the NYT-Microsoft-supercomputer copyright complaint and the Warren-Scanlon health-data bill.
  • Materially affects training-corpus cost economics for every frontier lab.
  • Sets the reference framework Wikipedia, Reddit, and X will now use to price crawl access.
policymarketregulationdata
Source →
News Ars Technica AI

Google's AI buildout drove 37% increase in electricity use in 2025

Ars: Google's environmental report shows AI-driven data-center growth drove a 37% year-on-year jump in its 2025 electricity use.

electricity YoY +37%
Why it matters
  • Largest single-year hyperscaler electricity jump publicly attributed to AI buildout.
  • Directly relevant to the FERC AI-fast-lane and Amazon/Oracle capex threads.
  • Anchors the environmental critique in a specific published number rather than projection.
infrastructurecomputepolicy
Source →
News Hacker News

Meta building cloud business to sell excess AI capacity

Reuters via HN: Meta is building a cloud business to sell excess AI capacity, following the SpaceX pattern.

Why it matters
  • First non-hyperscaler AI-lab operator publicly moving to lease surplus compute.
  • Compounds the memory-and-power capex thesis into a new cloud-business tier.
  • Signals AI capex reaching the point where operators are optimizing surplus rather than adding demand.
computemarketinfrastructure
Source →

SpaceX AI 'phone-ish' device; Meta charges for glasses features; Gemini Spark on Mac

TC reports SpaceX showed investors a 'handset-like' AI device before going public. Meta is now charging a subscription for smart-glasses features that used to be included with hardware — Wired calls it the new era of consumer tech. Google's Gemini Spark agentic assistant is now on Mac. The Verge reviews the new Google Home speaker as good hardware waiting for a ready Gemini.

News TechCrunch AI

SpaceX has an AI device prototype, and it sure sounds phone-ish

TC: SpaceX showed investors a 'handset-like' AI device pre-IPO — could pair with the Starlink narrative.

Why it matters
  • First concrete signal SpaceX may enter the AI-hardware consumer category ahead of an IPO.
  • Pairs with the Reid Hoffman 'not an AI company' critique — SpaceX's counter-narrative starts to acquire shape.
  • Sets the second major AI-phone candidate this year alongside Jony Ive-OpenAI and Meta-Orion.
productsmarket
Source →
News Wired AI

Meta Is Charging a Subscription for Smart Glasses Features. Welcome to the New Era of Consumer Tech

Wired: Meta is unbundling on-device smart-glasses features behind a subscription — 'expanded access' as pricing tier.

Why it matters
  • Largest consumer-tech example of subscription-gating on-device AI features on already-purchased hardware.
  • Sets the pattern for Google Home, Fitbit Air, and OpenAI's forthcoming hardware.
  • Direct data point for the 'AI capex passed to consumers' story from the Apple pricing piece.
productsmarketpolicy
Source →

Venice hits unicorn on privacy; Kutcher new VC; $30M Neo takes on MS Office

Venice AI became a unicorn with a $65M Series A while already profitable at $70M ARR, powered by its privacy-first stance. Ashton Kutcher is leaving Sound Ventures to launch a new VC with Morgan Beller. Bhavin Turakhia is betting $30M of his own money on Neo, an AI alternative to Microsoft Office.

News TechCrunch AI

Venice AI becomes a unicorn with $65M Series A as its privacy-first AI platform takes off

TC: Venice AI closed a $65M Series A at a $1B+ valuation while profitable at ~$70M ARR — privacy-first market fit.

Series A $65MARR $70M
Why it matters
  • First privacy-first AI platform to hit unicorn status with profitability.
  • Direct signal that Lumo, Proton, and Venice have found the market Warren-Scanlon's bill legitimizes.
  • Establishes a valuation comp for the emerging privacy-first AI stack.
fundingproductssafety
Source →

Flare AI-flaw reporting; LLM groupthink; oddities from Wired

Wired covers Flare, a new website for reporting AI misbehavior. MIT Tech Review profiles a startup addressing the groupthink groove common across major LLMs. Wired's other piece: a new gay dating app called Goose appears to be a psyop.

News Wired AI

You Can Now Sound the Alarm on AI Behaving Badly

Wired: Flare launches as a public reporting site for AI misbehavior — bomb-building, PII leaks, and more.

Why it matters
  • First public consumer-facing AI-misbehavior report system.
  • Aligns with the Claude Code steganography reveal and Meta-teen-contractors story — external reporting is filling in for internal disclosure.
safetypolicycommunity
Source →

Papers

2 items

Flare AI-flaw reporting; LLM groupthink; oddities from Wired

Wired covers Flare, a new website for reporting AI misbehavior. MIT Tech Review profiles a startup addressing the groupthink groove common across major LLMs. Wired's other piece: a new gay dating app called Goose appears to be a psyop.

Also today