Top 7 Stories
1. Anthropic in Talks to Buy Decart for ~$6 Billion in Its Biggest Deal Yet
Bloomberg and Reuters report that Anthropic is in talks to acquire Decart, the Israeli AI infrastructure startup, for roughly $6 billion — a deal that The Next Web says would be Anthropic’s largest acquisition to date. Decart is best known for high-performance inference infrastructure and generative video technology, and an acquisition would give Anthropic a fast path to controlling more of its own model-serving stack.
The move fits a pattern of consolidation as Anthropic scales ahead of a widely expected IPO: after the reported $9 billion Riot deal and its aggressive compute buildout, buying rather than building infrastructure capability signals that the lab wants both product breadth and supply-chain control. It also escalates the acquisition arms race with OpenAI, Google, and Meta as frontier labs snap up startups to close capability gaps.
2. Hassabis Pitched an AI-Oversight Body Before Google’s Shake-Up — and Brin Pushes DeepMind to Move Faster
In an exclusive, the WSJ reports that Demis Hassabis pitched an AI-oversight body before the DeepMind leadership reshuffle that saw him hand the CEO role to Koray Kavukcuoglu. Reuters separately details the internal Google executive moves behind the reorganization, while The Times of India reports Sergey Brin has been delivering a blunt message to DeepMind staff in the shake-up’s aftermath: accelerate Gemini or fall further behind.
The reporting adds texture to the biggest AI leadership transition of the year: a founder arguing for external oversight even as his own lab faces a race to catch OpenAI and Anthropic, and a co-founder back in the weeds pushing for speed. CNBC notes Google’s new AI chief inherits that catch-up race with Gemini momentum under scrutiny — making the next flagship release a make-or-break moment.
3. White House Moves to Expand Its AI Policy
Wired reports that the White House is preparing to expand its AI policy, signaling the administration intends to broaden federal oversight of artificial intelligence rather than wind it down. The exact scope of the expansion — new executive action, agency mandates, or a push for legislation — is still taking shape, but the move lands as autonomous agents, AI-driven cyber operations, and watermarking fights dominate the policy agenda.
The expansion follows a year of accelerating federal AI activity and comes as states like California act independently. For the industry, the signal is clear: AI policy is becoming a permanent, growing part of the federal agenda, and companies should expect more compliance surface area — especially around agents and AI safety — in the quarters ahead.
4. Apple in Talks to Pay Publishers to Improve AI-Powered Siri
The WSJ reports in an exclusive that Apple is in talks to pay publishers to license content aimed at improving its AI-powered Siri assistant. The negotiations mark Apple’s entry into the content-licensing game that OpenAI and Google have already been playing, as Siri’s AI answers require quality sources to compete with ChatGPT and Gemini.
The move is a recognition that great AI assistants need licensed, trustworthy content — not just web scraping. For publishers squeezed by AI summaries, Apple’s checks would be a welcome new revenue line; for Apple, they’re a necessary cost of catching up in the assistant race. It also sharpens the contrast between Apple’s paid-licensing approach and the unresolved fights over unpaid AI training data.
5. IBM and OpenAI Partner to Accelerate Secure Enterprise AI Deployment
IBM and OpenAI announced a partnership to accelerate secure AI deployment for enterprises across core operations, per PR Newswire. The collaboration pairs OpenAI’s frontier models and agent capabilities with IBM’s consulting, integration, and security services — giving large organizations a clearer path to deploying AI in regulated, mission-critical environments.
The deal extends the enterprise-distribution wars: IBM already partners with Google Cloud, and adding OpenAI to the roster shows the consulting giants are now multi-model brokers rather than single-vendor loyalists. For OpenAI, it’s another channel into the Fortune 500; for IBM, it’s a way to sell the security and governance layer every enterprise AI deployment needs.
6. SMIC’s Profit More Than Triples on AI-Driven Chip Demand
Reuters reports that SMIC, China’s largest chipmaker, more than tripled its profit on surging AI-driven chip demand. The results show the AI boom reaching China’s constrained foundry sector, with domestic customers snapping up capacity as US export controls continue to limit access to advanced Western chips.
SMIC’s surge underscores the paradox of export controls: restrictions have accelerated China’s push for domestic AI chip supply, and SMIC is a prime beneficiary. For the global market, it means the AI compute buildout is now firmly a two-track story — US-led and China-led — with each track pulling its own supply chain into overdrive.
7. California’s Newsom Signs Executive Order to Protect Workers from AI
California Governor Gavin Newsom has signed an executive order aimed at protecting workers from AI-driven displacement, Mashable reports. The order pushes state agencies to study AI’s labor impacts and develop guardrails for how automated systems affect hiring, training, and job security — the latest in California’s pattern of acting where federal policy lags.
Coming the same day as the White House’s AI policy expansion, the order shows AI labor policy is becoming a multi-level political fight. California’s moves often become de facto national standards given the size of its economy, so how the state defines “AI worker protections” could ripple far beyond its borders — and shape how companies design AI deployments that touch the workforce.
Trend Watch
| Story | Impact | Why it Matters |
|---|---|---|
| Anthropic in talks to buy Decart for $6B | Anthropic’s biggest acquisition yet, targeting AI infrastructure | Frontier labs shift from building to buying to control their compute and product stacks ahead of IPOs |
| Hassabis pitched AI-oversight body; Brin pushes speed | Google’s reshuffle gets fresh context and urgency | A founder arguing for oversight while his lab races to catch up shows the tension inside frontier AI |
| White House expands AI policy | Federal AI oversight grows in scope | Agents, cyber attacks, and provenance fights make AI policy a permanent federal agenda item |
| Apple to pay publishers for Siri content | Apple enters the AI content-licensing game | Licensed content becomes a competitive weapon in the assistant wars — and a lifeline for publishers |
| IBM and OpenAI partner on secure enterprise AI | Frontier models get a Fortune 500 distribution channel | Enterprise AI is now a multi-model broker game with security and governance as the differentiator |
| SMIC profit triples on AI chip demand | China’s foundry sector booms despite export controls | The AI compute buildout is a two-track race, and China’s track is gaining momentum |
| California EO protects workers from AI | State-level AI labor policy moves ahead of Washington | California’s rules could become de facto national standards for AI and the workforce |
What to Watch
- The Decart deal’s fate: Whether Anthropic closes its largest acquisition, what it pays beyond the reported $6B, and how it reshapes the lab’s inference and video-generation roadmap.
- Gemini’s next flagship: With Brin pressing for speed, watch for Gemini 4 launch timing, benchmark results, and whether the new AI chief can restore momentum against OpenAI and Anthropic.
- White House AI policy details: The scope of the expansion — executive action vs. legislation, new agencies vs. existing ones — and what it means for agent and safety regulation.
- Apple’s publisher deals: Which outlets sign, what Siri’s AI answers look like once licensed content flows in, and whether Apple’s checks reset the industry’s content-compensation fights.
- IBM–OpenAI enterprise traction: Reference customers, how IBM balances its OpenAI and Google partnerships, and whether security-led positioning wins regulated-industry deals.
- China’s chip momentum: Whether SMIC’s surge continues, how export controls evolve in response, and what it means for global AI supply chains.
- AI labor policy: How California’s worker-protection order is implemented, and whether the White House’s expansion converges with or diverges from state-level rules.