The Pentagon is launching a new task force to study the safe deployment of advanced AI models with hacking capabilities across NSA and Cyber Command missions. This critical initiative addresses rising concerns that private-sector AI tools could be leveraged to identify digital security vulnerabilities much faster than traditional hacking methods, necessitating a proactive and strategic military response.
A critical vulnerability in Fediverse software, including Mastodon and Misskey, has been identified, requiring urgent updates to prevent JSON-LD manipulation. Meanwhile, Firefox is phasing out asm.js optimizations in favor of WebAssembly. In the literary world, Granta and the Commonwealth Foundation are investigating AI controversies, while C++26 prepares to introduce new function wrappers like std::copyable_function.
Iranian-linked hackers are targeting US gas stations by exploiting unprotected fuel tank monitors, posing significant safety risks. Globally, Interpol has arrested hundreds in a major cybercrime crackdown across the MENA region. Meanwhile, privacy concerns grow as pacemaker data can be used as criminal evidence, and Dutch authorities investigate if stolen patient data remains a threat.
UN human rights experts have expanded their probe into Pornhub’s parent company to include tech giants like Google and Meta, targeting the infrastructure enabling abusive content. Meanwhile, Meghan Markle joined the WHO to warn that Big Tech poses a “dangerous bargain” for childhood safety. In cybersecurity, new NPM supply chain attacks are surging as global data breaches hit record levels.
Starbucks aims to save $2 billion by opening a new tech office in India, while Elon Musk seeks a supermajority at SpaceX to prevent his removal. US prosecutors are targeting over $1 billion in healthcare and pandemic-era fraud. Meanwhile, a whey protein shortage is looming due to surging demand, and Meta faces scrutiny after deleting a major Instagram account following a request from Kuwait.
The software industry faces growing technical debt from AI-generated code, while US job losses in AI-exposed roles dropped 0.2%. In transportation, Tesla Robotaxis were involved in two crashes during remote teleoperation. Meanwhile, one in seven Britons now turn to ChatGPT instead of a GP for medical advice, and Europe’s cloud sovereignty remains threatened by a heavy reliance on US-linked processors.
Anthropic is calling for stricter AI export controls to protect technological leadership. In cybersecurity, a Microsoft Exchange flaw is allowing attackers to use OWA inboxes for script launches. Google is testing a 5GB storage limit for new Gmail accounts, while X has pledged stronger content moderation in the UK. Additionally, the UK has ordered £1B in remote-control howitzers to replace aging systems.
The U.S. Department of Homeland Security has launched a new unit to re-evaluate green card holders for potential fraud, part of an intensified immigration enforcement effort. Meanwhile, Germany’s security agency has rejected Palantir software in favor of ChapsVision, citing privacy concerns. Additionally, growing privacy alarms over Meta Ray-Ban smart glasses are fueling calls for global regulation.
Critical security flaws in Meari-powered baby monitors expose 1.1 million families to potential surveillance. Geopolitically, US intelligence reveals Iran retains 70% of its missile stockpile, contradicting claims of military decline. In the UK, a relaunch of a voluntary digital ID scheme faces privacy backlash. Meanwhile, 50,000 Lake Tahoe residents face an energy crisis as utilities prioritize data centers.
Princeton University is reinstating in-person proctored exams to combat AI-driven cheating. In the markets, semiconductor stocks like Intel and Qualcomm have slumped following rising inflation and oil prices. Meanwhile, Google is updating Android with new spyware detection and a “Pause Point” to curb app addiction. Finally, Anthropic has launched 12 new AI plugins specifically for legal use.
