Washington becomes the gatekeeper for frontier AI

Saturday, June 27, 2026

The biggest AI story today isn't a model — it's who controls access to one. The US government has quietly become the chokepoint for frontier releases, even as markets convulsed over an OpenAI IPO delay and 'super bubble' warnings.

Trump Administration Becomes Gatekeeper for the Most Powerful New AI Models

Both OpenAI and Anthropic agreed to restrict their newest models at the Trump administration's request during a cybersecurity review, effectively handing Washington a veto over who can use frontier AI. It's a watershed: the government is now a gatekeeper for the most capable models, not just a regulator after the fact.

Read the brief →

OpenAI Hands the Government a Veto Over Who Gets Its Newest AI

OpenAI launched GPT-5.6 on June 26 but locked most users out, citing a US government request to limit access. The release makes the new federal gatekeeping concrete — the company's most powerful model shipped and was simultaneously walled off.

Read the brief →

US Lifts Block on Anthropic's Most Powerful AI, Clearing Mythos 5 for 100+ Institutions

On the flip side, the government lifted its hold on Anthropic's Claude Mythos 5, clearing it for 100+ US institutions. It shows the review process cuts both ways — and that approval, not just restriction, now flows through Washington.

Read the brief →

Anthropic Accuses Alibaba of 28.8 Million-Query Attack on Claude

Anthropic accused Alibaba of running the largest-known 'distillation' campaign against Claude — 28.8 million queries over 44 days to clone its capabilities. The charge sharpens US-China AI tensions and raises the question of how labs can defend their models from being copied through the API.

Read the brief →

AI Stocks Melt Down as 'Super Bubble' Warnings Spread Across Global Markets

A broad sell-off in AI stocks dragged Wall Street toward a losing week as 'super bubble' warnings spread through global markets. After a year of one-way euphoria, investors are openly debating whether the rally has outrun the fundamentals.

Read the brief →

OpenAI Weighs Delaying Its IPO to 2027 — and Markets Wobble

A report that OpenAI may push its IPO to 2027 sent a ripple through global equities — a sign of how much market sentiment now hinges on the company. The mere rumor of a delay was enough to move tech shares worldwide.

Read the brief →

OpenAI Spices Up the Chip Wars With 'Jalapeño,' Built to Lean Less on Nvidia

OpenAI unveiled Jalapeño, its first custom inference chip built with Broadcom to power ChatGPT and future models. It's a direct bid to lean less on Nvidia and control the economics of actually running AI at scale.

Read the brief →

Micron Soars to Record High as AI Memory Demand Quiets the Doubters

Micron blew past forecasts and hit a record high on surging AI memory demand, a rare bright spot amid the week's bubble jitters. Strong memory pricing is hard evidence that AI buildout spending remains very real on the hardware side.

Read the brief →

AI and Chip Stocks Rebound, Pulling Wall Street Off Its Lows

AI and chip stocks bounced back late in the week, pulling Wall Street off its lows after several brutal sessions. The whipsaw underscores how jumpy the AI trade has become — sharp drops and sharp recoveries in the same week.

Read the brief →

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