01 Claude paid subscriptions have more than doubled as user estimates vary widely

Anthropic told TechCrunch that paid Claude subscriptions have more than doubled so far this year, a tangible sign of growing commercial traction for its consumer offerings. External estimates of total consumer users vary considerably — TechCrunch cited figures ranging from about 18 million to 30 million — but Anthropic has not published a definitive total user count.

Those divergent numbers reflect the challenge of measuring AI product adoption across free and paid tiers, partner integrations, and embedded uses. Anthropic’s spokesperson framed the subscription growth as a concrete metric of paying‑customer demand even while broader user estimates remain 'all over the map.'

For competitors and customers, the key takeaway is the split between rapid paid‑tier growth and uncertainty about aggregate reach: investors and partners can point to subscription momentum, while market observers still debate how many people use Claude across all channels.

Takeaways
  • Anthropic reports paid‑subscription growth has more than doubled year‑to‑date.
  • Independent estimates of total consumer users range roughly between 18 million and 30 million.
  • Subscription metrics offer a clearer commercial signal than headline user estimates, which remain inconsistent.

02 xAI keeps shrinking: Elon Musk’s last reported co‑founder has left

TechCrunch reports that the most recently remaining co‑founders of xAI have exited, leaving only two of the original 11 still at the company. The publication notes that all but two of the co‑founders had departed before this week, and the newest departure further reduces the early founding team.

Founder attrition of this scale is notable for a high‑profile, founder‑led AI startup because it reshapes internal leadership and institutional knowledge. The departures do not, on their own, reveal the company’s product roadmap or technical progress, but they are a concrete signal about changes in the organization’s composition.

For observers tracking the competitive AI landscape, xAI’s co‑founder exodus is a reminder that early teams often evolve quickly — and that management and talent changes are part of the risk profile for startups building advanced models.

Takeaways
  • All but two of xAI’s original 11 co‑founders had left before this week, per TechCrunch.
  • The latest reported departure removes another founding leader from the company’s roster.
  • High founder turnover changes governance and continuity but is not by itself evidence about model or product performance.

03 Court grants Anthropic a preliminary injunction as U.S. AI policy shifts and scrutiny widens

A federal judge granted Anthropic a preliminary injunction that temporarily blocks a Pentagon ban while the company’s lawsuit proceeds, The Verge reports. The ruling follows a weeks‑long standoff over the Department of Defense’s designation of Anthropic as a supply‑chain risk, and it preserves Anthropic’s ability to pursue contracts pending a full legal resolution.

The injunction arrives amid simultaneous shifts in U.S. AI policy and personnel: The Verge also reported that David Sacks, who had been advising the White House as a special advisor on AI and crypto, is no longer serving in that official role. His departure changes one senior external voice that had been linked to the administration’s AI policy engagement.

Separately, legislative scrutiny of AI infrastructure is increasing: TechCrunch covered a bipartisan request from Senators Josh Hawley and Elizabeth Warren asking the Energy Information Administration for more data on how data centers use power and the effect on the electrical grid. That push for power‑use transparency underscores rising congressional interest in the physical and policy footprint of AI services.

Takeaways
  • A judge issued a preliminary injunction that pauses the Pentagon’s ban on Anthropic while litigation continues.
  • David Sacks is no longer a White House special advisor on AI and crypto, removing a prominent industry advocate from that role.
  • Senators Hawley and Warren asked the Energy Information Administration for more details on data‑center power use, signaling increased oversight of AI infrastructure.
Briefs

What moved around the edges

04

Epismo launches CLI to package repeatable human‑AI workflows

Epismo released an npm‑based CLI that saves and shares multi‑step agent workflows; the founder says the tool exports chains of markdown tasks and works with Claude Code and other agents, with 380+ downloads reported after launch.

Hacker News AI
05

Researchers identify a large cavity in near‑Earth space that may reduce cosmic‑ray exposure

A report summarized by 404 Media describes a giant void in galactic cosmic rays created by Earth’s magnetic field, which researchers say could help reduce radiation exposure for astronauts in some regions.

404 Media AI
06

Conntour raises $7M to build natural‑language search for security video

Conntour secured $7 million from General Catalyst and Y Combinator to build an AI search engine that lets security teams query camera feeds in natural language to find people, objects, or events.

TechCrunch AI
07

Google updates Gemini 3.1 Flash Live to improve audio capabilities

Google published a Gemini 3.1 Flash Live post describing model improvements aimed at making audio AI more natural and reliable in live settings.

Google AI Blog
08

Google expands Search Live to more countries and languages

The Verge reports that Google’s Search Live — a voice‑and‑camera AI search assistant — is now available in over 200 countries and territories and supports dozens more languages as part of a broader rollout.

The Verge AI

Subscribe to AI Digest

A compact morning briefing built from primary AI sources.