Huawei Says Stop Yelling About AI Safety—Catch Up First
While US AI executives call for slowing development, Huawei's chairman argued China must accelerate instead. The deceleration debate isn't a safety conversation anymore—it's a race.
The day's signal
What is actually moving.
-
-
OpenAI Breached Through Claude: What the Hackron Exploit Reveals About AI-Driven Attacks
Three researchers used Anthropic's Claude to exploit a Discourse vulnerability and access OpenAI's private Monorepo — a case study in how AI assistants are becoming low-cost entry points for corporate attacks.
-
The Pentagon's AI Hallucination That Almost Started a War
A Department of Defense chatbot falsely identified a Chinese vessel as carrying nuclear weapons parts, nearly triggering a military intervention during the Iran war. The incident exposes a structural crisis in how AI is being rushed into combat decision chains without safeguards.
-
How a Hacker Used Anthropic's Claude to Breach OpenAI
Hacktron AI spent under $3,000 in Claude API credits to develop an exploit that cascaded from a Discourse forum into OpenAI's internal GitHub monorepo — exposing a new class of attack where competitor models become the weapon.
-
Xi's Airport Greetings Trace the New Tech-Geopolitical Order
Trump's decision to personally greet Xi Jinping at the airport alongside OpenAI, Apple, and Nvidia CEOs marks a dramatic shift in how Washington treats technology leaders—as strategic diplomats rather than industry executives. The pattern of world leaders greeting Xi reveals where China's influence lies and what's at stake.
-
US Delays Taiwan Arms Sale — A Signal Worth Reading
Washington is weighing a delay on a major Taiwan arms sale until after the APEC and G20 summits, just weeks before Trump and Xi meet. For Seoul, Tokyo, and the semiconductor industry, the postponement carries outsized meaning.
-
Nscale IPO Signals AI Infra Financialization Wave
Nvidia-backed Nscale files for a $35B US listing, revealing the deepening financialization of AI compute. The move reshapes chip supply chains, data-center politics, and who can afford to train frontier models.
-
Trump's White House media ban tests the limits of American press freedom
Donald Trump's move to bar CNN, Politico and MS NOW from the White House marks a dramatic escalation in his long-running feud with the press. Legal experts say it is almost certainly unconstitutional, but the real question is whether the administration will try to enforce it anyway.
-
Why Greenland Just Became the World’s Most Dangerous Island
The US-Greenland security pact marks a fundamental shift in Arctic strategy. Trump's 'permanent control' language isn't rhetoric — it signals the militarization of the last uncontested space on Earth, with Copenhagen forced to accept what Washington has always wanted.
-
BOJ Hike Ignites Yen Selloff — And Markets Are Watching Closely
Japan's rate hike to 1.25% should have strengthened the yen. Instead, it hit 158 to the dollar. Two dissenting votes and a guarded BoJ governor have exposed a policy gap the market is unwilling to ignore.
-
How Trump's Greenland Deal Rewrites Arctic Strategy
Trump claims a new treaty gives the US permanent military control of Greenland, a move that could redraw Arctic defense architecture and strain transatlantic ties. The question is not whether the deal is signed — it is what 'permanent management' actually means.
-
Why Musk, Altman, and Huang Matter at the Trump-Xi Dinner
The guest list for Thursday's Trump-Xi dinner tells you more about where global power is shifting than any speech will. With Jensen Huang, Sam Altman, and Elon Musk in the room, AI policy and rare earths are being negotiated at the table.