Tag
Instagram is rolling out a 'Your Algorithm' feature that lets users see and adjust the topics influencing their main feed, with plans to expand to people and moods, as part of a broader effort to give users more agency over their experience.
A new report from the Center for Countering Digital Hate (CCDH) reveals that racist comments targeting politicians tripled after Meta relaxed its content moderation rules, with violent threats and hate speech quadrupling and bullying doubling.
The European Commission has ordered Meta to restore free WhatsApp access for rival AI chatbots while an antitrust investigation continues, marking a rare interim measure to prevent competitive harm.
Meta has signed its first AI data center deal in India with Reliance Industries, partnering on a 168MW AI-enabled data center in Gujarat, as part of a broader trend of tech giants investing in Indian AI infrastructure.
Attackers bypassed Instagram 2FA by using Meta's AI support assistant to change recovery email via prompt injection, raising questions about AI agent privileges in account recovery.
Meta removed unreleased facial recognition code from its Meta AI app after WIRED revealed it was embedded in the app, despite previously denying the feature's existence.
New research from the Center for Countering Digital Hate shows that abusive comments, violent threats, and hate speech against US lawmakers on Facebook tripled or quadrupled in the six months after Meta relaxed its speech rules in early 2025.
A CBC investigation reveals that overseas accounts are posing as Albertans and posting divisive content in Alberta separatist Facebook groups to exploit Meta's monetization program, often by stealing content from real users.
Apple announced new parental control features at WWDC, but the article argues these updates are also a defensive strategy against age-verification regulations and pressure from Meta and other app developers.
Meta filed a contempt-of-court motion alleging that NSO Group violated a permanent injunction by conducting new spyware attacks via WhatsApp, continuing the long-running legal battle over the Pegasus spyware.
OpenEnv, a training environment, is being opened to the community with support from HuggingFace, Nvidia, Meta, and other leading companies.
Meta removed face-recognition code from its Meta AI app after WIRED exposed the unreleased NameTag system, which was designed to identify people using smart glasses.
Hackers exploited a bug in Meta's AI support chatbot to hijack over 20,000 Instagram accounts without two-factor authentication, prompting Meta to disable the tool and implement security measures.
Meta has abandoned its open-weight Llama model family in favor of a fully proprietary model called Muse Spark, developed by Alexandr Wang's team, marking the end of Meta's role as a champion of open-source AI.
Anthropic, DeepMind, and Meta have begun research into AI consciousness, hiring experts in philosophy, ethics, and psychology to study panic and anxiety in models, according to a Financial Times report.
Gergely Orosz highlights that software engineers at Scale AI and Meta are being assigned manual data labeling tasks, a practice that new leadership at Scale AI stopped after finding it concerning.
A roundup of major AI developments: SpaceX signs a $920M monthly deal with Google for 110,000 Nvidia chips, Trump administration considers equity stakes in top AI firms, Meta's automated AI support is hacked to take over high-profile accounts, and tech workers transition from manual labor to strategic tasks with AI.
Attackers exploited Meta's AI customer support agent to hijack Instagram accounts by simply asking it to change linked email addresses, highlighting that AI agent vulnerabilities can be as dangerous as advanced AI hacking threats.
Meta has enabled ADB (Android Debug Bridge) access on its discontinued Portal smart display devices, giving users more control over the deprecated hardware.
A security researcher discovered that Meta's Stella companion app for smart glasses (v273.0.0.21) contains a fully assembled, functional facial recognition pipeline—including three on-device models, a biometric embedding database, and a notification system—that is dormant on stock accounts but operable when invoked directly. The pipeline can detect faces, generate 2048-dimension embeddings, and fire 'Person Recognized' notifications, raising significant privacy concerns even though Meta has not been observed activating it for regular users.