Tech News — 2026-06-14#

Story of the Day#

Anthropic’s flagship “Mythos” AI models are at the center of a major geopolitical storm after reports surfaced that a group linked to China may have accessed the system. The security breach reportedly triggered swift White House export restrictions, prompting global leaders to openly warn about the systemic risks of relying too heavily on a consolidated American AI ecosystem.

Top Stories#

State Attorneys General Launch Sweeping Investigation Into OpenAI · The New York Times A coalition of state attorneys general has officially opened an investigation into OpenAI’s core business practices. Regulators are heavily scrutinizing the AI giant’s handling of consumer data, its advertising activities, and how its products impact the safety of minors.

Meta Forced to Unwind $2B Manus Deal After Beijing Intervention · TechCrunch Meta is backing out of its $2 billion acquisition of Manus after the Chinese government explicitly demanded the deal be reversed. The forced dismantling of the acquisition signals a severe, direct escalation of Beijing’s interference in global technology M&A.

US and Japan Pour $1B into ‘Genesis’ AI Platform for Science · Slashdot The US Department of Energy is attempting to build a “national operating system for science” by integrating 17 national labs, supercomputers, and AI agents into a unified research platform called the Genesis Mission. Japan has signed on as the first international partner, bringing the initial joint investment to $1 billion in an effort to maintain a strategic lead over China in quantum tech, fusion, and biotechnology.

UK Prepares Under-16 Social Media and Chatbot Ban · Bloomberg The United Kingdom is poised to ban children under 16 from using social media platforms and AI chatbots, mirroring similarly aggressive legislation recently proposed in Australia. Culture Secretary Lisa Nandy acknowledged the ban isn’t a “silver bullet,” but described it as a necessary structural change to protect child safety online.

Meta’s $14B Pivot to Proprietary AI is Failing to Impress Developers · Slashdot Despite spending $14 billion and hiring Alexandr Wang to build its “Muse Spark” model, Meta’s sudden shift away from its open-source Llama approach is facing heavy skepticism from the developer community. With its virtual reality division bleeding over $80 billion, Wall Street is growing impatient, demanding that Mark Zuckerberg deliver tangible AI monetization rather than just closed-garden ad features.

Also Worth Knowing#

• FBI Builds Fake City to Simulate Cyberattacks (The Verge): The bureau has constructed a 22,000-square-foot mock town in Alabama, complete with a hackable 200-server data center, to run physical and digital cyber warfare training.
• First UK Police Officer Investigated for AI Evidence Forgery (Slashdot): A British police officer is under criminal investigation for allegedly perverting the course of justice by using artificial intelligence to fabricate evidential material.
• Bitcoin Down 50% in 11 Months (Slashdot): Bitcoin has cratered to around $64,000, shedding nearly half of its value since hitting a record high above $123,000 in July 2025.
• Machine Intelligence Causes 50% Crypto Token Wipeout (Bloomberg): In a stark warning for the security industry, a crypto token collapsed after an AI exposed a fatal flaw that human cryptographers had missed for years.
• Java’s Looming Upgrade Crisis (Slashdot): Enterprise IT faces a massive modernization bottleneck as four Long-Term Support versions of Java (8, 11, 17, and 21) are all scheduled to hit end-of-support in a compressed three-year window between 2029 and 2032.