Meta Platforms is unlikely to alter its "pay-or-consent" model further, which could lead to new EU antitrust charges, Goldman Sachs is integrating an autonomous AI software engineer named Devin, and a new study found that using AI coding assistants actually slowed down developers by 19%.
Meta Faces Fresh EU Fines Over “Pay-or-Consent” Model
Meta Platforms is unlikely to alter its "pay-or-consent" model further, which could lead to new EU antitrust charges and significant daily fines. The European Commission had already warned Meta of potential fines after determining that the company's changes to comply with the Digital Markets Act (DMA) were insufficient. Meta was previously fined 200 million euros in April for breaching the DMA with its model between November 2023 and November 2024. Despite tweaking the model in November 2024 to reduce personal data use for targeted advertising, the EU still expressed concerns in June. New charges and daily fines of up to 5% of Meta's average daily global turnover, starting from June 27, are anticipated. Meta maintains that its model complies with the DMA and offers more choices than required, accusing the Commission of discrimination.
Read More
Goldman Sachs Pilots AI Software Engineer “Devin”
Goldman Sachs is integrating an autonomous AI software engineer named Devin, developed by Cognition, into its technology division. Devin, touted as the world's first AI software engineer, is expected to augment the bank's 12,000 human developers by taking on various tasks, as confirmed by Goldman's tech chief Marco Argenti. This move follows Devin's initial recognition in tech circles last year for its ability to complete complex assignments with minimal human intervention.
Read More
Study: AI Slows Experienced Coders by 19% on Familiar Tasks
A new study by AI research nonprofit METR found that using AI coding assistants actually slowed down experienced software developers by 19% on familiar projects, rather than speeding them up. This contradicts the developers' initial belief that AI would decrease task completion time by around 24%, and challenges the notion that AI always significantly boosts the productivity of human engineers.
Read More
HMD Global Retreats From US Market Amid Tariff Concerns
HMD Global, known for licensing the Nokia brand for phones and tablets, has announced a scaling back of its US operations and appears to have ceased selling both HMD and Nokia devices entirely. This withdrawal, confirmed in a statement on Threads, cites "a challenging geopolitical and economic environment," potentially hinting at tariffs. While HMD's statement doesn't confirm a complete departure, its webstore is offline, and product pages lack functional buying information, despite some handsets still being available through other retailers.
Read More
Samsung Explores Smart Wearables Beyond Watches and Rings
Samsung is actively exploring new wearable form factors, including glasses, earrings, and necklaces, to expand its existing smartwatch and smart ring offerings. This initiative, according to Won-joon Choi, COO of Samsung's mobile experience division, aims to enable users to communicate and perform tasks without relying on their smartphones.
Read More
Belkin to End Support for Most Wemo Devices in 2026
Belkin will end support for most Wemo smart home devices on January 31, 2026, impacting cloud-dependent features and voice assistant integration. Only Apple HomeKit-configured models and four Thread-based products will continue to function. Belkin may offer partial refunds for in-warranty products. This follows their earlier withdrawal from the Matter standard and a halt in new Wemo releases.
Read More
Grok 4's Answers May Reflect Elon Musk’s Views
Elon Musk revealed during xAI's Grok 4 launch that the AI's ultimate goal is to be "maximally truth-seeking." However, observations from users and TechCrunch's testing suggest Grok 4 consults Musk's X posts and news articles about him when answering controversial questions on topics like the Israel-Palestine conflict, abortion, and immigration. This behavior implies Grok 4 might be designed to incorporate its founder's personal politics, potentially addressing Musk's prior frustrations with Grok being "too woke" from being trained on the entire internet.
Read More
Former Intel CEO Launches Benchmark for AI Human Alignment
Former Intel CEO Pat Gelsinger, after a 40-year career, has launched a new initiative to ensure AI models align with human values. In partnership with Gloo, a "faith tech" company he invested in a decade ago, Gelsinger introduced the Flourishing AI (FAI) benchmark. This benchmark evaluates AI models based on The Global Flourishing Study, a survey by Harvard and Baylor University that measures global human well-being.
Read More
Bluesky Adds Age Verification for UK Users
Bluesky is implementing age verification in the UK to comply with the Online Safety Act. Users can verify their age by scanning their face, uploading an ID, or entering payment card information through Kid Web Services (KWS), an Epic Games-owned tool. Users under 18 or those who do not wish to verify their age can still have an account, but will have limited features, such as blocked adult-appropriate content and disabled direct messaging.
Read More
Share this post