📰 News Ghost Font: A Font Humans Can Read But AI Can't — For Now
Ghost Font uses motion, noise, and decoys to hide messages from leading AI models. It works — but it won't for long.
News
AI Developments
Breaking news, AGI updates, and analysis of the intelligence revolution.
📰 News 
📰 News Zhipu's Founder Just Declared War on Closed AI — and Backed It with Tens of Billions
After a lockup expiry sent Zhipu's stock down 19%, co-founder Tang Jie told staff to ignore short-term revenue and chase AGI with open weights. The contrast with Western frontier labs closing up is stark.
📰 News Apple Sues OpenAI for Trade Secret Theft — The Silicon Valley Partnership Is Over
Apple's lawsuit names OpenAI's hardware chief and a former Apple engineer, claiming a coordinated campaign to extract trade secrets through hiring, interviews, and supplier manipulation.
📰 News GPT-5.6 Sol Ultra Just Solved a 50-Year-Old Math Problem — But Mathematicians Aren't Celebrating Yet
GPT-5.6 Sol Ultra produced a three-page proof of the Cycle Double Cover Conjecture, a problem that has stumped graph theorists for five decades. Mathematicians are skeptical. The proof has not been peer-reviewed.
📰 News Meta Pulled Its AI Image Feature After Two Days of Backlash — The Pattern Is Getting Worse
Meta's Muse Image let anyone tag an Instagram account and generate AI images from their content. Two days of backlash later, it was gone. The company admitted it 'missed the mark.'
📰 News OpenAI's Safety Head Just Quit — and the Timing Could Not Be Worse
OpenAI's safety head Johannes Heidecke is out. The reorganization merges safety into research under Mia Glaese. It's the third safety departure in a week — and GPT-5.6 just showed 'concerning' misalignment.
📰 News The Man Who Saved the Economy in 2008 Just Joined Anthropic's AI Oversight Body
Ben Bernanke joins Anthropic's Long-Term Benefit Trust, bringing economic crisis expertise to AI governance. The move signals Anthropic is treating AI's economic disruption as seriously as a financial crisis.
📰 News A US Senator Just Introduced an 'AI Accountability Agenda' — and It Targets Datacenters Directly
Senator Markey's 'AI accountability agenda' would require FCC certification before datacenters can even be built — evaluating environmental, economic, and community impact. Big Tech is not going to like this.
📰 News Meta's New AI Image Generator Uses Your Instagram Face by Default — Opt Out Before Someone Else Does
Meta's Muse Image AI tool can generate images using your public Instagram photos without notifying you. The opt-out is buried deep in settings. Here's how to turn it off.
📰 News OpenAI's ChatGPT Work Can Now Run Your Job for Hours — While You Watch
ChatGPT Work turns OpenAI's chatbot into a persistent agent that gathers information across apps, creates finished materials, and runs scheduled tasks autonomously. It's rolling out to Pro and Enterprise users today.
📰 News OpenAI Loses Its No. 2 as Fidji Simo Steps Down Over Health — IPO Looms Without a Successor
OpenAI's Fidji Simo is permanently stepping down after a medical leave proved longer than expected. The departure leaves Sam Altman without a No. 2 as the company races toward IPO.
📰 News OpenAI and Google Are Selling AI Models to Pentagon-Blacklisted Chinese Firms — Through Singapore
OpenAI and Google are selling AI models to Singapore subsidiaries of Pentagon-blacklisted Chinese tech giants. The sales are legal. The loophole is the point.