Tech News — 2026-04-13#

Story of the Day#

OpenAI CEO Sam Altman was targeted in a second violent attack at his San Francisco home on Sunday morning, following an earlier firebombing incident on Friday. The escalating violence, which has resulted in the arrest of two suspects for a drive-by shooting and a 20-year-old anti-AI activist from Texas for the Molotov cocktail attack, marks a terrifying and tangible shift in the temperature of the global AI debate.

Top Stories#

Anthropic Withholds “Mythos” AI Over Severe Cybersecurity Risks Anthropic has built its most capable frontier model yet, Claude Mythos Preview, but is refusing to release it to the public because it is simply too good at hacking. In internal tests, the model autonomously discovered and exploited zero-day vulnerabilities in every major operating system and web browser, including a 27-year-old bug in OpenBSD and a 16-year-old vulnerability in FFmpeg. Instead of a public rollout, Anthropic launched “Project Glasswing,” giving only a select consortium of tech giants—like Apple, Google, and Microsoft—access to the model to patch critical software vulnerabilities before bad actors can exploit them.

The 2026 AI Index Proves AI is Sprinting While Regulation Crawls Stanford’s annual AI Index dropped today, revealing that the US is crushing the global compute race with an estimated 5,427 data centers—over ten times as many as any other nation. The report highlights a technology that is evolving faster than current benchmarks can measure, driving massive capital expenditure while the hardware supply chain remains perilously dependent on a single Taiwanese foundry, TSMC. Meanwhile, the public remains deeply anxious, with 52% reporting they are nervous about AI’s impact, as lawmakers around the world scramble to pass meaningful oversight.

Roblox Forces Mandatory Age Tiers to Tackle Child Safety Crisis Under intense legal scrutiny, Roblox is drastically overhauling its platform by automatically sorting users into strict age-gated tiers: Kids (ages 5-8), Select (ages 9-15), and regular accounts (16+). By June, unverified users will be locked into a restricted “Kids” experience with chat disabled by default and only minimal-maturity games accessible. It is a massive structural shift for the gaming giant, but critics are already questioning whether automated age verification and AI moderation will actually be enough to keep enterprising children out of restricted, mature spaces.

Meta Faces Massive Pushback Over Smart Glasses Facial Recognition A coalition of over 70 civil liberties groups is demanding that Meta kill its planned “Name Tag” facial recognition feature for Ray-Ban smart glasses before it even launches. The feature uses an AI assistant to pull up the identity of bystanders in the wearer’s field of view, prompting severe warnings that the tech will silently arm stalkers, scammers, and federal agents. Internal documents suggest Meta hoped to slide the launch through while advocacy groups were distracted by a dynamic political environment, but this preemptive strike proves the privacy battle is just beginning.

Slate Auto Bags $650M for an Actually Affordable EV Truck Jeff Bezos-backed EV startup Slate Auto just raised a fresh $650 million to begin delivering a bare-bones electric pickup truck aimed at the mid-$20,000s. In a market obsessed with six-figure luxury EVs, Slate is targeting the low-end utility segment and aiming to undercut the Ford Maverick XL as the cheapest pickup in America. The Series C round was led by TWG Global, giving the startup a massive cash injection as it pushes to get the first vehicles delivered later this year.

Maine Moves to Become First US State to Ban New Data Centers Lawmakers in Maine have approved a bill that will block the construction of new data centers in the state until November 2027. The aggressive moratorium aims to establish guardrails to prevent AI infrastructure build-outs from driving up local energy prices for residents. The bill now awaits final signature from Governor Janet Mills, who had previously pushed for geographic exemptions that the state House ultimately rejected by a wide margin.

Apple Developing Display-Free AI Glasses to Rival Meta Apple is quietly designing display-free AI smart glasses to compete directly with Meta’s Ray-Bans, aiming for a 2027 release. The wearables will feature high-end acetate frames, distinctive oval cameras, and deep iPhone integration. Apple’s strategy leans heavily on a significantly upgraded Siri coming in iOS 27, using computer vision to interpret a user’s surroundings without the friction or bulk of a full AR headset.

Also Worth Knowing#

• Xbox Game Pass Has Become “Too Expensive”: In a leaked internal memo, Microsoft’s new Xbox chief Asha Sharma admitted that recent price hikes made Game Pass “too expensive for players” and hinted at evolving the subscription into a more flexible model.
• IBM Pays $17M to Settle DOJ’s Anti-DEI Lawsuit: IBM paid the US government $17 million to resolve claims from the Trump administration that its corporate diversity policies violated the False Claims Act by discriminating against applicants.
• Booking.com Confirms Major Customer Data Breach: The travel giant admitted hackers accessed customer reservations, exposing names, contact details, and private accommodation messages to potential phishing scams.
• Linux 7.0 Officially Released: Linus Torvalds released the new Linux 7.0 kernel, dropping the experimental label from Rust support and advancing the transition to initramfs for early-userspace booting.
• Google Shoehorns Rust Into Pixel 10 Modem: To combat remote code execution vulnerabilities in baseband processors, Google integrated a Rust-based component into the legacy C/C++ codebase of the Pixel 10’s cellular modem.