Tech News
← Home  ·  All topics

Ai

5382 GoKawiil briefs on this topic · RSS for ai

IGF AI-disclosure rule sparks 150-comment debate over Big Walk entry

The Independent Games Festival this year required every entrant to declare any AI use, a stricter standard than Steam's player-facing disclosure rule. This led many indie developers to disclose AI use for the first time, with the game Big Walk becoming the focus of controversy after its developers said LLMs were used for less than 0.1% of its code to fix late-production errors. A judge for the festival described the resulting forum debate, which grew to roughly 150 comments, as unusually heated even by IGF standards.

YC-backed Tinfoil hiring applied AI security engineer in San Francisco

Tinfoil, a Y Combinator-backed security research lab building a private, verifiable AI cloud, is hiring an Applied AI Security Engineer for its San Francisco office, requiring in-person work at least three days a week. The role involves guiding enterprise and AI lab customers through technical discovery and deployment, helping them verify Tinfoil's security guarantees, and feeding customer insights back to product and engineering teams. Tinfoil counts DuckDuckGo, Mozilla, Google DeepMind, Trail of Bits, Red Hat and Canonical among its customers or research collaborators.

CoreWeave surges 16% premarket after 168% revenue jump, lifted 2026 guidance

CoreWeave shares jumped more than 16% in premarket trading after the company reported third-quarter revenue up 168% year-over-year and raised its 2026 guidance, lifting other neocloud stocks along with it. The rally comes even as CoreWeave remains down 65% from its May peak, having lost over 40% of its value in the past month alone. Nvidia shares also rose about 3% premarket amid reports its board hired law firm Wachtell Lipton Rosen & Katz to investigate CEO Jensen Huang over possible security leaks.

Madras High Court orders probe into alleged AI-generated books at Tamil Nadu library

The Madras High Court has directed Tamil Nadu's anti-corruption agency to investigate claims that a newly opened public library in Coimbatore purchased hundreds of suspected AI-generated books, and to freeze accounts linked to the supplying publisher. The probe follows allegations by novelist Perumal Murugan that nearly 1,800 Tamil-language books at the Periyar Arivulagam library, built at a cost of about $34m, showed signs of dubious authorship, inflated prices and questionable ISBNs.

Google launches Tabletop, an AI tool for building shared family and group apps

Google introduced Tabletop, an experimental AI platform that lets users generate custom collaborative tools—like savings trackers, chore charts, trip itineraries, dinner polls, expense splitters and tournament brackets—by describing what they need or choosing a template. Once created, a space can be shared via invite link so multiple people can use and edit it simultaneously, similar to a Google Doc. The beta is currently available only to US users aged 18 and over through a waitlist, according to Google Labs design and engineering lead Roman Nurik.

Fiber-optic shortage tied to AI data centers hits Ukraine's drone supply chain

A BlueBird Tech representative told Ars Technica that fiber-optic cable prices have risen six to ten times over the past two years, partly due to AI data center demand consuming global supply. Ukraine reportedly acquired only 72 miles of Chinese fiber-optic cable in August 2025, compared to Russia's 328,000 miles that same month, with Russia importing an estimated 37 million miles across 2025.

Meta brings Muse AI agent companion app to Wear OS and Apple Watch

Meta announced that its Muse AI agent app is now available on Wear OS and Apple Watch, letting users interact via text, dictation, or voice chat directly from their wrist. The smartwatch app functions as a remote control, allowing users to approve or check on tasks, but the agent itself still runs on a paired phone rather than natively on the watch.

Gmail expands Gemini thread summaries and writing aid to more Workspace users

Google is rolling out two existing Gemini-powered Gmail features — AI Overview conversation summaries and Help Me Write — to a broader set of Workspace accounts, including school and work users who previously lacked access. The features were already available to personal Gmail accounts and some Workspace tiers. Access requires users to be 18 or older with Gmail smart features turned on, and the rollout is expected to reach eligible accounts within 15 days.

Ex-Facebook employee Dave Goldblatt launches site tracking AI's real-world wins and failures

Dave Goldblatt, who previously worked at Facebook, has created a website called Because of AI that documents concrete examples of artificial intelligence succeeding and failing in practice. The project aims to catalog real-world outcomes rather than hype or doom narratives surrounding the technology.

Stingray Marine Solutions scales AI laser drones to 3,500 units across Norway's salmon farms

Stingray Marine Solutions, based in Oslo, reports deploying over 3,500 submarine AI laser robots to protect more than 112 million farmed salmon, up roughly 1,000 units from last year. The drones use AI to detect and zap sea lice with lasers, an approach the company calls 'optical delousing' that now covers 20-25% of Norway's farmed salmon, according to its latest annual report. The 250-person company says the system operates continuously without chemical treatments.

CrowdStrike exec says AI-hacking panic is overblown, serves other interests

Adam Meyers, head of counter-adversary operations at CrowdStrike, described facing intense questioning over a hack of six South Korean banks during a recent trip, which he uses as a jumping-off point to argue that fears over AI-powered hacking are often exaggerated. He and other cybersecurity professionals say such alarm can distort how companies prioritize defenses and distracts from more mundane, proven attack methods.

LinkedIn tests 'Frequent Viewer' tag showing who repeatedly visits your profile

LinkedIn is piloting a feature that flags users who have viewed a Premium subscriber's profile multiple times, labeling them as 'Frequent Viewers.' This builds on the existing Who's Viewed Your Profile tool, which already lets Premium users see a year's worth of visitors. LinkedIn confirmed the test but did not provide further details on its rollout or availability.