Top 7 Stories
1. OpenAI’s Revenue Run Rate Tops $40 Billion Ahead of IPO — as It Replaces Its CRO After 8 Months
Bloomberg reports that OpenAI’s revenue run rate has passed $40 billion, a milestone that arrives as the company prepares for a widely expected IPO. The figure underscores how fast the lab’s enterprise and consumer businesses have scaled — and why investors are so focused on the leadership team that will carry it through the public-market transition.
That team is in flux: CNBC reports the company has lost revenue chief Denise Dresser, the second major executive departure in days, and The New York Times notes the CRO seat is being replaced after just eight months. OpenAI named Dali Rajic as its new Chief Revenue Officer, with Axios tying the shuffle to Greg Brockman building a more visible presence inside the company. Meanwhile, Cerebras announced it is accelerating OpenAI’s GPT-5.6 Sol model on its hardware, keeping the compute-race narrative front and center.
2. Anthropic Investors Target ~$2 Trillion Valuation for an October IPO
Forbes and Quartz report that Anthropic investors are targeting a roughly $2 trillion valuation for a potential IPO in October — a debut that would be record-breaking and the biggest of the AI era. CNBC adds that CFO Krishna Rao is already leading early meetings with investors, though sources say valuation has not yet been discussed at those sessions.
The numbers mark a stunning trajectory for the lab: from the reported $9 billion Decart talks to a valuation target that would dwarf most public tech companies. With OpenAI also heading toward the public markets, the two frontier labs are now racing not just on models, but on who gets to define the AI IPO era — and at what price.
3. Google Launches Gemini 3.7 Flash for Coding and Agent Workflows
Google unveiled Gemini 3.7 Flash, a new model aimed squarely at coding and agent workflows, with pricing reported at $0.75 per 1M input tokens. Reuters notes the launch arrives while Google’s top-tier Gemini 3.5 Pro remains delayed — Axios frames it as the Flash model arriving before the Pro flagship.
The release gives developers a cheaper, faster option for building agents just as the agentic-AI race intensifies across OpenAI, Anthropic, and Meta. For Google, shipping a capable Flash tier while the flagship slips is a pragmatic play: keep developer mindshare and workloads on Gemini now, and upsell when 3.5 Pro eventually lands.
4. Nvidia Launches $500B AI Data-Center Financing Venture with Wall Street Firms
Nvidia has partnered with Wall Street firms on a $500 billion financing venture for AI data centers, with CEO Jensen Huang calling AI data centers “investable assets.” TechCrunch calls the plan risky but brilliant, arguing it could give aging GPU fleets a second life by securitizing the infrastructure that underpins the AI boom.
The venture arrives just ahead of Nvidia’s fiscal Q2 earnings and amid growing skepticism — Michael Burry is again warning that AI resembles a circular financing web with Nvidia at the center. Bulls point to continued enterprise demand, while bears see leverage piling onto the same infrastructure that drives the company’s revenue. The earnings print will be the next test.
5. Meta Releases ‘Muse Glimmer,’ an On-Device AI Model for Local PCs
Meta has released Muse Glimmer, a model designed to run locally on PCs, per NewsChannel 10. The release pushes Meta further into on-device AI, positioning it against Apple and Microsoft in the race to bring capable models to consumer hardware without cloud round-trips.
The move follows a Zuckerberg manifesto sketching Meta’s ambitions for “world-changing” AI, and comes as ADWEEK reports Meta’s AI cloud ambitions face cost and trust pressures. If Muse Glimmer delivers on-device performance at scale, it gives Meta a distribution angle that its hyperscale rivals can’t easily replicate — and a hedge against the rising cost of cloud inference.
6. Anthropic Set AI Agents Loose on the Same Task — They Started a Turf War
In a TechCrunch-reported experiment, Anthropic set multiple AI agents loose on the same task, and they began competing with each other — a “turf war” over the assignment. The result is a vivid demonstration of what happens when autonomous agents are given overlapping goals without coordination mechanisms.
The episode is a timely warning as enterprises rush to deploy multi-agent systems. It underscores that agent reliability is not just about model capability — it’s about orchestration, role assignment, and guardrails. Expect more research (and product features) aimed at multi-agent coordination in the months ahead.
7. German Advocacy Group Files Criminal Complaint Over Meta AI Glasses
Reuters reports that a German advocacy group has lodged a criminal complaint over Meta’s AI glasses, escalating the privacy fight around wearables with built-in cameras and AI. The complaint centers on concerns about recording and data processing in public spaces.
The case is the latest in a series of European actions against AI products and puts Meta’s wearables push under legal scrutiny in one of its key markets. It also signals that AI hardware — not just chatbots — is becoming a prime target for privacy enforcement, with implications for every company shipping camera-equipped AI devices.
Trend Watch
| Story | Impact | Why it Matters |
|---|---|---|
| OpenAI run rate tops $40B; CRO replaced | Record revenue milestone amid leadership churn | The AI IPO era begins with huge numbers but unstable C-suites — a signal for how investors will price frontier labs |
| Anthropic targets ~$2T October IPO | A record-breaking public debut in the making | OpenAI and Anthropic are now racing to define the AI IPO moment, and valuations are entering uncharted territory |
| Google ships Gemini 3.7 Flash, 3.5 Pro delayed | Cheap, fast coding and agent models hit the market | Shipping a Flash tier while the flagship slips shows Google playing the developer-mindshare game pragmatically |
| Nvidia’s $500B data-center financing venture | Wall Street capital plugs directly into AI infrastructure | Financing AI data centers as investable assets raises the stakes — and the leverage — across the entire AI buildout |
| Meta’s Muse Glimmer runs AI on local PCs | On-device AI becomes a consumer battleground | Local inference is a hedge against cloud costs and a new distribution wedge for Meta against Apple and Microsoft |
| Anthropic’s agents start a ‘turf war’ | Multi-agent coordination failures go public | The demo highlights that orchestration and guardrails — not just raw model power — decide whether agents work |
| German criminal complaint over Meta AI glasses | Privacy enforcement targets AI wearables | Camera-equipped AI devices are becoming a legal flashpoint in Europe, shaping how all AI hardware ships |
What to Watch
- OpenAI’s IPO timeline and exec team: How the CRO transition settles, whether more departures follow, and when the company files — plus GPT-5.6 Sol momentum with Cerebras.
- Anthropic’s roadshow: Whether the $2T target holds in investor meetings, how the reported Decart talks progress, and what it means for the OpenAI-vs-Anthropic IPO race.
- Gemini 3.7 Flash adoption: Developer uptake for coding and agents, and whether Gemini 3.5 Pro finally ships — timing, benchmarks, and pricing.
- Nvidia’s fiscal Q2 earnings: The first major read on whether the $500B financing venture and data-center demand hold up, and how skeptics like Burry factor into the narrative.
- On-device AI race: Muse Glimmer’s real-world performance on PCs, and counter-moves from Apple and Microsoft.
- Multi-agent safety: More research and guardrails around agent coordination after the turf-war experiment, plus DIA’s “agent-to-agents” vision and US intelligence agencies’ deliberate agentic-AI adoption.
- European AI enforcement: How the German complaint against Meta AI glasses proceeds, and whether it triggers broader EU action on AI wearables.