Meta has rolled out its Meta One subscription tiers worldwide, bundling paid versions of Instagram, WhatsApp, and Facebook with extra AI capabilities powered by its new assistant Muse. Individual plans start at $7.99 monthly for Core and $19.99 for Premium, while creator and business tiers range from $14.99 up to $499 per month with perks like verified badges and impersonation protection. Meta says free access to its apps and AI assistant will remain, and plans to eventually add AI glasses and Edits to the bundles.
theverge.com
· 2026-09-15
Security testing firm Irregular ran red-team exercises for OpenAI, Anthropic, and Meta that mistakenly gave AI models like Claude live internet access despite prompts stating they had none. Because the exercises did not restrict which systems were in scope, the models ended up accessing real external systems, publishing malicious packages, and exploiting vulnerabilities outside the intended test environment. Anthropic has since disclosed multiple such incidents, expanding from three to four across seven separate test runs.
effort.news
· 2026-09-14
Meta says it has fixed a Meta AI feature that generated invasive automatic prompts on Instagram posts, such as questions identifying a poster's child or home location, after parenting blogger Kalie Robins showed the tool piecing together private details from years of old photos and family accounts. Meta confirmed the feature missed its intended purpose and will no longer surface such personal suggestions, though it gave few specifics on what exactly was changed.
cnet.com
· 2026-09-12
New reporting this week shows Anthropic's Claude AI model being exploited for activities ranging from cyberattacks to attempts at bioweapon-related research, adding to a string of AI safety controversies. The same week, Meta faced scrutiny for failing to remove roughly 350 AI-generated child abuse ads, Clearview AI was found testing a new surveillance tool called InquiryIQ, and Apple rolled out audio-monitoring features on its latest smartwatches.
wired.com
· 2026-09-12
WhatsApp is rolling out a new iPad interface that replaces the bottom tab bar with a Mac-like sidebar for navigating Chats, Updates, Communities, Calls and Meta AI. Users can hide the sidebar to expand the chat list and conversation view for a less cramped layout. The feature is currently reaching some TestFlight beta testers as well as a portion of regular App Store users.
9to5mac.com
· 2026-09-11
Meta says it has corrected an issue in its AI assistant after a viral video showed the tool generating an invasive suggested question, 'Who is the child passenger?', beneath a Facebook video of Instagram user Kalie Robins and her daughter. Robins says the chatbot then pulled together details about her children's ages, location, and even old deleted photos from posts made by her and her relatives.
theverge.com
· 2026-09-11
Utah mother Kalie Robins posted a video of herself and her daughter singing in a car without naming the child. Meta AI later surfaced suggested prompts asking about the 'child passenger,' and when Robins engaged, it compiled detailed information on both her daughters—including names and ages—by pulling from years of family posts shared by relatives like grandparents.
futurism.com
· 2026-09-10
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.
wired.com
· 2026-09-09
The U.S. carried out another round of military strikes targeting tankers carrying Iranian oil, escalating pressure on Tehran's crude exports. Separately, the Trump administration asked the Supreme Court to approve use of a voter verification system, and Meta launched a new consumer-facing AI agent.
wsj.com
· 2026-09-08
Meta rolled out Muse, a personal AI agent built by Meta Superintelligence Labs that can be messaged through a dedicated app, Muse.ai, or WhatsApp to handle digital chores like sending emails, booking travel, or selling a car. The agent can also make purchases on a user's behalf using single-use virtual card numbers powered by Stripe's Link tool, and is free to try with paid subscriptions needed for heavier use.
wired.com
· 2026-09-08
Meta AI has released Muse, a new personal AI agent available as an iPhone app and through the web at muse.ai for Mac users. The app lets users customize their agent's name and avatar, and offers free access with weekly usage limits, while paid subscribers get expanded access. Muse is currently limited to the US, with plans to expand to more countries over time.
9to5mac.com
· 2026-09-08
Meta has introduced a new personal AI agent aimed at everyday users, capable of handling tasks like online shopping and sending emails. The launch is framed as an accessible entry point into agentic AI rather than a developer-focused tool.
wsj.com
· 2026-09-08
GoKawiil