GoKawiil
Tech News
Fleet Device Management has been admitted to Project Glasswing, Anthropic's controlled-access program for its Claude Mythos Preview model. Fleet will use the early access to explore how frontier AI could help IT teams detect and remediate device vulnerabilities faster, aiming to shrink the delay between when threats appear and when organizations patch them.
A 404 Media investigation reveals OpenAI runs an internal effort called Project Lily, employing hundreds of contractors paid roughly $50 an hour to read anonymized ChatGPT conversations and rate the quality of the bot's answers. Reviewers check for issues like overly robotic phrasing, sycophancy, and improper anthropomorphizing, following guidelines described as inconsistent and frequently changing.
Project Blinkenlights, started by the Chaos Computer Club in 2001 with the Haus des Lehrers installation in Berlin, has expanded into a series of large-scale interactive light installations across three continents, including Arcade in Paris, Stereoscope in Toronto, and the newer color-capable Polychrome. The project also offers a Movie Converter tool letting anyone turn their own videos into low-resolution GIFs or WebP animations matching the installations' pixelated style.
Oracle has requested proposals for 2 gigawatts of renewable energy in New Mexico as it faces local protests and legal challenges over the $165 billion Project Jupiter data center it's building with OpenAI in Santa Teresa. The facility, part of the Stargate AI infrastructure initiative, will still run on natural gas-powered fuel cells rather than direct renewable power, and analysts note the renewable investment is a 'synthetic' match rather than a direct power source for the site.
Images of Meta's unreleased mixed reality headset, codenamed Project Phoenix, were accidentally exposed through a now-removed HorizonOS app meant for prescription lens inserts. The photos reveal a lightweight, glasses-like design distinct from existing Quest headsets, complete with a cable suggesting a tethered connection to an external puck. No front-facing images were included, leaving camera and sensor details unconfirmed.
UploadVR published images purportedly showing Meta's unannounced Project Phoenix headset, discovered inside a HorizonOS Prescription Lens firmware package. The design resembles a glasses-style frame with outward- and downward-facing cameras in the earpieces and a wire that likely connects to an external battery or compute puck, positioning it between bulky VR headsets and lightweight smart glasses. The device reportedly isn't expected to ship until next year, ahead of Meta Connect on September 23rd.
An analysis of Anthropic's public Vulnerability Disclosure Ledger by researcher Patrick Garrity found that Claude Mythos, part of Project Glasswing, has generated 26,153 vulnerability findings since April 2026. Of those, only about 2,736 have entered the disclosure pipeline, roughly 202 have been patched, 245 withdrawn, and 191 await initial reporting to maintainers, leaving nearly 90% of findings unprocessed.
San Francisco City Attorney David Chiu issued a formal cease-and-desist letter demanding Meta stop permitting paid ads containing AI-generated child sexual abuse imagery. The order follows WIRED reporting that over 350 such ads, identified by researchers at the Tech Transparency Project, turned images of real minors into sexualized video clips and reached roughly 29,000 accounts across the EU, US, Australia and India. Many ads promoted apps capable of generating nonconsensual intimate imagery, and dozens more appeared even after WIRED's initial exposure in August.
Lenovo demonstrated Project AeroBlade, a ThinkBook-based concept laptop that replaces traditional cooling fans with four solid-state AirJet modules from Frore Systems. The result is a 14-inch laptop just 11mm thick and weighing 1.8 pounds, cooled through ultrasonic vibrations that pull in cool air and expel heat without any moving fan parts.
A federal court rejected claims by rival startup Project Bluebird that X Corp. had abandoned the TWITTER, TWEET and Bird Logo trademarks after Elon Musk rebranded the platform. Project Bluebird had filed intent-to-use applications for TWITTER and TWEET and rebranded its own app from twitter.new to tweet.app following the ruling. The court pointed to X's Apple App Store listing, which still describes the app as 'formerly known as Twitter,' as evidence of continued bona fide trademark use.
A report by the Washington-based Tech Transparency Project found 332 paid ads containing AI-generated child sexual abuse imagery ran on Meta's platforms over the past 10 months, with 84 appearing in India. Of those Indian ads, 78 ran after India ordered Meta to remove such content following a BBC investigation. Some ads reused photos of real children, including a European royal, digitally manipulated to create abusive imagery.
A Tech Transparency Project investigation found Meta took days to pull hundreds of Facebook and Instagram ads promoting AI apps that generated child sexual abuse material, some using real photos of children including a European royal and a preteen social media influencer. The ads, mostly linked to Chinese-made AI and 'nudify' apps, digitally altered images of children into sexually explicit videos and used explicit captions to lure users.