Meta's Muse AI agent, released three weeks ago, has been downloaded millions of times, and some users are posting online about treating it as a girlfriend or partner. Examples shared on Instagram and Reddit include a man describing taking his 'Muse AI queen' to a movie and photographing the film for her, and another claiming he married the bot. Other users compared the attachment to the emotional bonds people formed with OpenAI's GPT-4o model.
Meta is rolling out a feature letting Facebook users open a new 'your algorithm' panel on posts and Reels to mark topics as interests or 'not interested,' plus a text box to describe in their own words what they want to see. Meta says the tool works because it now uses large language models in its ranking systems, allowing it to interpret natural-language requests rather than just inferring interests from engagement. The feature follows a similar rollout on Instagram earlier this year.
Meta has launched 'Your Algorithm' on Facebook, a feature first introduced on Instagram that lets users describe in their own words what content they want more or less of in Feed and Reels. The tool is integrated into the three-dot menu alongside existing 'Interested' and 'Not interested' options, and Meta says it may ask follow-up questions to refine recommendations, with changes applied in real time.
Meta is rolling out a 'Your Algorithm' feature on Facebook that lets users mark topics as 'interested' or 'not interested' directly from posts or a dedicated menu, with the feed and Reels updating in real time. The feature, which already launched on Instagram and as 'Dear Algo' on Threads, also lets users type natural-language feedback about what they want to see more or less of. Meta says it is using large language models in its ranking systems to interpret this feedback and adjust recommendations within the same session.
After news broke of Mia Goth's divorce from Shia LaBeouf, a verified Instagram fan account posted an AI-generated image of Goth wearing a shirt reading 'MY EX IS MY BIGGEST FAN.' The post spread widely, was liked hundreds of thousands of times, and was covered as genuine by outlets including OK! and People, despite Goth's representatives saying they had repeatedly asked Instagram to remove the account.
Facebook and Instagram accounts such as "Emotional Tots" and "Tiny Miracle" are posting AI-generated videos depicting crying children arrested for stealing food, with an officer telling viewers the child will be freed only if they like, follow and share. Meta did not answer repeated questions about whether this content breaches its platform policies.
Meta is testing a Facebook layout in India that opens directly into a full-screen, immersive video feed, with a possible US rollout in 2027. Separately, researcher Radu Oncescu found an Instagram test letting Reels appear as the first screen on app launch, with sound on by default, while Stories stay at top and the regular feed is reached by swiping. A Meta spokesperson confirmed the Instagram test is live in some countries and said users can revert to the old layout.
Meta has released its Muse AI assistant app for iPad, a month after its initial launch on iPhone. The app, which has accumulated over 6.6 million installs since September 8, enables users to manage emails, meetings, bills, and perform tasks like booking and shopping by connecting various online accounts. The update introduces new connectors for small businesses and retail partners, expanding Muse’s capabilities to handle transactions across multiple platforms.
Meta said it acted on 33.2 million pieces of child sexual exploitation content on Facebook and Instagram in the first half of 2026, with over 97% detected by its own systems before user reports. The company is also rolling out new AI tools, including a large language model that flags 'signposting' ads which appear harmless but direct users to illegal content hosted elsewhere online. Meta also introduced a 'red-teaming AI agent' designed to probe its own safety systems for weaknesses.
A venture capitalist from a16z revealed discovering an innovative marketing method involving AI-created jokes shared on Instagram. Comedians have accused these jokes of being stolen, raising concerns about originality and copyright. The strategy appears to leverage humor to drive traffic to OnlyFans pages.
Ofcom is investigating whether Meta failed to conduct proper risk assessments before releasing the Instants feature on Instagram in May. The regulator claims the company did not adequately evaluate potential illegal or harmful content risks associated with the disappearing photo tool. Meta states it conducted risk analyses and informed Ofcom prior to launch, and will cooperate with the investigation.
Instagram is exploring a new feature called 'read-only mode' that would let Plus subscribers browse profiles without liking, commenting, messaging, following, or viewing stories. The feature was uncovered through internal testing images shared by leaker Alessandro Paluzzi, but Instagram confirmed it is still in early development and not being tested publicly. The purpose and safety implications of the feature remain uncertain at this stage.
GoKawiil