GoKawiil
Tech News
Meta is expanding its AI-based age estimation technology on Instagram worldwide to detect teen users even if they list an adult birthday, placing them into stricter Teen Account protections. The system analyzes profile content like captions, comments and photos for contextual and visual cues, including school mentions and physical appearance. Meta says it will extend similar tools to Facebook and later Threads this year.
The media industry is again investing heavily in social video production, years after a mid-2010s push collapsed when Facebook was found to have inflated video view counts, leading to mass layoffs of writers and video staff alike. This new wave is driven by changes to social platforms' algorithms favoring video and the rise of TikTok and YouTube, with YouTube now commanding nearly 16 percent of TV viewership.
Meta has indefinitely delayed changes to its commercial terms that would have required Facebook Business account holders to allow extensive AI training on any external websites linked from their pages. A Meta spokesperson told Reuters the company is pausing the changes to keep engaging with stakeholders, and that existing terms remain in effect for now.
Mark Cuban said on the Pivot podcast that Meta continues running Facebook ads using an AI-generated version of his face and voice to sell a fake diabetes cure, even as elderly victims email him asking if the product is real. He cited internal Meta documents reported by Reuters showing the company projected about $16 billion, roughly 10% of its 2024 revenue, from ads tied to scams and banned products, with Meta's own safety team estimating its platforms play a role in a third of successful U.S. scams.
Facebook accounts are posting AI-generated, near-photorealistic videos depicting male coaches inappropriately touching young female gymnasts, dancers and yoga practitioners, some appearing to be teenagers. Users who reported the content say Meta has declined to remove it, even though the clips depict groping, kissing and touching framed as normal interactions between coaches and athletes. A safe-sport advocate quoted in the report calls the material abuse content that sexualizes and normalizes real power imbalances in youth sports.
Meta is expanding parental notifications about self-harm content to Facebook accounts and conversations with its Meta AI assistant, after first launching the feature on Instagram. Parents will be alerted if a supervised teen repeatedly searches for self-harm-related content in a short period, or if their chats with Meta AI suggest risk of suicide or self-harm. Alerts arrive via text, email, WhatsApp or push notification, without revealing specific search terms or chat content, and point parents to helpline resources.
Meta spokesperson Andy Stone said the company may postpone implementation of controversial new business terms beyond their current November 20 start date, pending a full consultation with publishers and industry groups. The terms, already delayed once from an original October 30 date, would let Meta use automated scraping tools on the websites of businesses that use Facebook. Meta has not changed the actual wording of the terms despite the planned consultation.
A CNET writer used Meta's AI assistant Muse to automatically search Facebook Marketplace for an ottoman, air purifier and humidifier within set price ranges and locations. Instead of a one-time search, Muse set up a recurring task that rescans listings every four hours and surfaces matching items, including a specific air purifier brand the writer wanted.
Meta is testing restrictions that cap how many links Facebook pages can post per month unless they pay for a Meta One subscription starting around $50. Greenville, Texas communications manager Daniel Starks says the city's library page was capped at two links monthly, despite Meta saying nonprofits and government pages should be exempt. Other verified nonprofit pages report the same restrictions, prompting some organizations to stop posting links directly and instead place them in comments.
Santa Fe District Court Judge Francis J. Mathew ordered Meta to pay a $25 billion civil penalty, following a September jury verdict that found Facebook violated New Mexico's consumer protection laws more than 43 million times. The state's lawsuit centered on the Cambridge Analytica scandal, alleging Facebook gave third-party apps unlawful access to user data, mishandled hate speech, and profited from misinformation. Meta has said it will appeal the ruling.
Meta has reached an agreement allowing Mark Zuckerberg to sell shares without losing control of the company, aimed at keeping him in California if a proposed billionaire tax passes. The measure, a one-time 5% levy on residents worth over $1bn, goes before California voters on 3 November. Zuckerberg had warned Meta's board he was considering moving his primary residence out of the state.
Meta Platforms disclosed in its third-quarter earnings report that it plans to split its stock by introducing a new class of non-voting shares. The company said the change is set to take effect in 2027.