Meta Faces Social Media's "Big Tobacco" Moment as 29 States Take the Company to Trial in California
Meta's biggest legal battle yet begins Tuesday in an Oakland federal courtroom, where a coalition of 29 state attorneys general will argue the company deliberately designed Facebook and Instagram to hook children. The numbers on the table are staggering: Meta's own attorneys have said damages could reach $1.4 trillion, while the states' lawyers call $200 billion more likely. And the trial opens less than two weeks after Meta lost a similar case in New Mexico that will cost it nearly $1 billion.
Sony's Digital Monopoly Under Fire: Physical Games End in 2028 and Fans Are Asking Who Really Owns Their Library
Sony is pulling the plug on physical games beyond January 2028, and the backlash has cracked open a much bigger question: when every purchase is digital and the PlayStation Store is the only shop in town, what do players actually own? The answer, buried in the fine print, is a revocable license. And a recent dispute with an indie studio shows how much power sits on Sony's side of that arrangement.
ChatGPT Can Now Remember Everything You Do on Your Mac, and It's Off by Default for a Reason
OpenAI began rolling out Computer History on Thursday, a feature that lets ChatGPT and its coding agent Codex build a timeline of what users do across apps and websites on their Mac. The company is betting that recording clicks, typing and app switches, rather than screenshots, will spare it the backlash that sank Microsoft's similar Recall feature.
Twitch Streamers Revolt as Amazon Trains AI on Their Content by Default: "Nobody Would Opt In"
Twitch confirmed on Wednesday that parent company Amazon can use streamers' content to train its generative AI models, and every user is enrolled by default. The confirmation came wrapped inside what the platform framed as good news: a new setting that lets creators opt out. The community's answer was immediate and furious.
Nearly 100 Unknown Super Mario Bros. Cartridges Surface in a Wisconsin Storage Unit, Stunning Collectors
A box of almost 100 never-before-seen Super Mario Bros. and Duck Hunt cartridges has surfaced at a small game shop in Waukesha, Wisconsin, and the collecting world is scrambling to explain it. The variant was completely unknown until now, the cartridges were once destined to be melted down for gold, and several have already earned perfect grades never before given to a Nintendo Entertainment System game.
iOS 27 Hints at 'Apple Reference Image' Photo Authentication
A privacy disclosure buried in iOS 27 beta 5 has revealed details about a feature Apple is calling "Reference Image" -- a built-in photo authentication system that would allow iPhone cameras to certify a photo's origin at the exact moment of capture. The feature is not live yet, but the documentation is specific enough to give a clear picture of how it would work and who it is meant to serve.
TechnologyAugust 10, 2026 Zuckerberg Predicts Personal AI Agents for Everyone: What It Means
On Meta's Q2 2026 earnings call, held July 29, 2026, Mark Zuckerberg made a prediction that cut through the usual quarterly noise: within five years, billions of people will have personal AI agents that understand their goals, work around the clock, and manage significant parts of their lives. Finances, health, relationships, household logistics, all handled by an AI that knows you well enough to act on your behalf.
It's an ambitious vision. It also landed in the middle of a complicated financial moment for Meta.
TechnologyAugust 10, 2026 Google Deletes User's Purchased Lord of the Rings Trilogy From Google Play Library, Refuses Refund
A Reddit user who paid for the extended editions of Peter Jackson's The Lord of the Rings trilogy on Google Play says the films disappeared from their library without warning, and that Google has refused to issue a refund.
The user, posting under the handle "ugoindownsaka1," purchased all three extended editions in 2022 through Google Play, which has since been rebranded as Google TV. At some point after the purchase, the titles were quietly pulled from the platform's catalog as part of what appears to be a routine content rotation. The films did not simply become unavailable to new buyers; they were removed from the user's existing library entirely.
TechnologyAugust 10, 2026 OpenAI Admits Its Next AI Model Might Be Too Dangerous to Control, But Won't Show the Proof
OpenAI is sitting on evidence that its next major AI model could cross a threshold the company itself defined as "Critical" for cybersecurity risk. The catch: they won't release the data to prove it, or disprove it.
That tension is drawing serious scrutiny from AI safety researchers and independent evaluators. For anyone paying attention to where this technology is heading, the details matter.