OpenAI Offers Washington a 5% Stake in Itself

Saturday, July 4, 2026

The AI giants are redrawing their relationships with governments and suppliers today—OpenAI dangling equity to Washington, Anthropic hunting an escape from Nvidia. Here's what matters.

OpenAI Floats Giving Washington a 5% Stake in Itself

OpenAI has floated handing the US government a 5% ownership stake in the company, per the FT and confirmed by multiple outlets. It's an extraordinary bid to align with Washington amid regulatory scrutiny—and raises thorny questions about state influence over frontier AI.

Read the brief →

Anthropic Joins the AI Chip Race, Courting Samsung to Escape Nvidia

Anthropic has begun early work on a custom AI chip and held preliminary talks with Samsung to manufacture it, according to The Information. It's the clearest sign yet that top labs want to break their dependence on Nvidia's costly, supply-constrained GPUs.

Read the brief →

US Reverses Course, Lifts Export Ban on Anthropic's Most Powerful AI Models

The US lifted export controls on Anthropic's most powerful models, restoring global access. The reversal signals a friendlier federal posture toward American AI leaders competing abroad—and pairs notably with Anthropic's other moves today.

Read the brief →

Nvidia Will Trade Chip Access for a Slice of Startups' Future Revenue

Nvidia unveiled an optional financing model letting AI startups access GPUs now and pay later, partly by sharing future revenue. It deepens Nvidia's grip on the ecosystem, turning scarce compute into equity-like leverage over the startups that depend on it.

Read the brief →

Meta's Secret 'Watermelon' AI Now Rivals OpenAI's Best, Its Chief Says

Meta's superintelligence chief Alexandr Wang told staff an upcoming model codenamed 'Watermelon' now rivals OpenAI's best, per Business Insider. If true, it validates Meta's costly talent-and-compute spree and reshapes the frontier race just as rivals negotiate with governments.

Read the brief →

Anthropic Moves to Shut the Back Doors Chinese Firms Used to Reach Claude

Anthropic is tightening enforcement to block Chinese firms from reaching Claude through indirect routes, the FT reports. It shows how much manual policing export restrictions still require—and how leaky 'access' controls remain in practice.

Read the brief →

AI Stocks Wobble Even as the Dow Storms to a New Record

AI-related stocks sold off even as the Dow jumped 600 points to a record, per Investopedia. The split-screen suggests investors are growing choosier about AI exposure rather than buying the theme wholesale.

Read the brief →

Takeda Bets Up to $600M on Insilico's AI to Hunt New Drugs

Takeda partnered with Insilico Medicine on AI-driven drug discovery in a deal worth up to $600 million. It's another marker that big pharma increasingly treats AI as core R&D infrastructure, not experiment.

Read the brief →

The AI Briefing The day's most important AI, chip & tech stories — one email each morning. Free.