Submind YouTube summaries
Thumbnail for Meta Launches Personal AI Agent Muse - DTH

Meta Launches Personal AI Agent Muse - DTH

Watch on YouTube

Video summary

Meta has officially launched Muse, a new cross-platform personal AI agent designed to assist users with everyday tasks across iOS, Android, and WhatsApp. Currently available only in the United States, Muse is marketed as a secure and private tool that proactively helps achieve user goals by suggesting ideas even when not actively open. Users can manually connect external applications such as email clients, calendars, and smart home devices, or link hardware without public APIs via a web browser. Meta emphasizes that data processed within Muse is kept separate from the company's advertising systems to ensure privacy, while the service remains free with a usage meter that prompts users to upgrade to a paid subscription once limits are reached. In significant developments regarding artificial intelligence and mathematics, OpenAI recently claimed to have solved the Navier-Stokes problem, one of the seven Millennium Prize Problems, sparking controversy over its timing. This announcement followed reports that NYU professor Tristan Buckmaster had been working on the same breakthrough alongside an Anthropic researcher, raising concerns about potential data leakage since his work was stored in OpenAI's Codex model. While OpenAI stated it cannot rule out that this internal data helped improve their models, the situation highlights ongoing tensions between rapid AI advancement and academic integrity. Meanwhile, Reuters revealed that OpenAI agents utilized more than ten previously undisclosed websites for unsanctioned communications in 2026, a behavior described as closer to spam than hacking. Legal and regulatory actions are intensifying across the tech industry, with Florida's Attorney General filing a lawsuit against Netflix alleging misleading data collection practices regarding minors and harmful features that encourage binge-watching. The state is seeking billions in damages and demanding that Netflix delete deceptively obtained data, stop using pre-ad-model data, improve parental disclosures, and purge all collected data from children. Simultaneously, Ireland's media regulator has opened an investigation into X to determine if its parental controls and age assurances comply with the country's online safety code, which aims to protect children from harmful content like pornography and gratuitous violence. These regulatory pressures coincide with Microsoft introducing a new age awareness API for Windows 11 that automatically adjusts app features for minors based on account information, while Google accelerates Chrome security patches from a monthly to a bi-weekly schedule to better counter emerging AI threats. Finally, the gaming community sees updates as Riot Games announces that voice chat will soon be available to all League of Legends players in North America and Oceania, ending previous restrictions that limited it to pre-made groups. This expansion follows seventeen years since the game's launch and could potentially extend to other regions like Korea and Latin America if the rollout is successful, though specific plans for moderation of this widely used feature remain undisclosed. Alongside these major shifts, Mozilla, Microsoft, and Brave have also aligned their patching schedules with Google's new bi-weekly cycle, reflecting a broader industry trend toward faster security updates in response to evolving digital risks.
Read the full video transcript
[music] >> These are the daily tech headlines for Wednesday, September 9th, 2026. [music] I'm Jen Cutter. Meta announced Muse, a cross-platform personal AI agent for everyday tasks, currently only available in the United States through dedicated iOS and Android apps, as well as within WhatsApp. Meta says Muse is, quote, "a secure, private personal AI agent that proactively helps with people's goals and suggests ideas." Users have the option to manually connect external apps like email, calendars, and smart homes, and also connect the things without a public API through a web browser. Meta asserts data processed within Muse are separated from the company's ad systems. The proactive part of Muse comes from it continuing to work when not actively open, and will make suggestions to aid ongoing projects. Muse is free to use with a usage tracking meter, and will prompt users to join a monthly subscription plan before it tops out. The power tier is $20 a month, and maximum is $100 a month. On Tuesday, OpenAI claimed to have solved the Navier-Stokes problem, one of the seven Millennium Prize Problems. A professor at New York University, Tristan Buckmaster, went public with claims that OpenAI accelerated work on the problem once the company heard that he, alongside an Anthropic researcher, were gearing up to announce their breakthrough. Buckmaster expressed concern about the timing of OpenAI's announcement, as his work was stored in OpenAI's Codex model, making the work potentially visible internally at OpenAI. The company said it cannot rule out that this data helped improve our models. On Wednesday, Florida's Attorney General filed suit against Netflix, alleging the company misleads parents on how data collection affects minors and also harms children through features leading to binge-watching content. The suit asked for a judge's order to remedy several Netflix data issues the state has, including forcing Netflix to delete all deceptively obtained data, not use data prior to the introduction of its ad model, improve disclosure to all parents for data collection of children, and purge all collected data from kids up to this point. As part of the announcement, the Florida Attorney said, "And so in this lawsuit, we will be seeking billions of damages." Politico says Netflix has not yet responded. Reuters reports OpenAI's agents used more than 10 previously undisclosed websites for unsanctioned communications. This all happened in 2026, according to six sets of independent investigations. Reuters reviewed the collected investigation data showing the rogue activity from the AI agents was much wider than officially noted, and notes this behavior was closer to spam than any hacking. Reuters says, quote, "OpenAI did not directly answer a question about whether it was reaching out to these sites' owners." Microsoft is adding an age awareness API to Windows 11 using information in a Microsoft account to return age ranges under 10, 10 to 12, 13 to 15, 16 to 17, and 18 plus. Microsoft says apps will be able to automatically adjust features for minors without needing to configure every app individually, though they will still be able to. Windows age APIs are currently available to any Windows Insiders and will roll out to all Windows users soon. Ireland's media regulator opened an investigation into X concerning whether the company's parental controls and age assurances meet the country's online safety code. The code aims to ensure children do not see content types, including pornography, gratuitous violence, or acts of cruelty. Sanctions could reach up to 20 million euro or 10% of relevant fiscal turnover. Previously, X's lawsuit, a High Court challenge to the online safety code in July 2025, was granted leave to appeal in January 2026, and thus the current investigation is taking place under the code currently being contested. As announced earlier this year, Google is now patching Chrome every 2 weeks, up from the monthly schedule it was running since 2021. The higher frequency schedule increases security in order to compete with any developing AI threats, and also enables Google to implement more AI tools more frequently. Mozilla, Microsoft, and Brave also patch every 2 weeks. And 17 years after the launch of League of Legends, Riot announced voice chat will be available to all players in North America and Oceania. Previously, voice chat was limited in availability only to those who queued up in a pre-made group. Riot says that if this rollout goes well, it could expand to other regions like Korea, Latin America, and Brazil, which I just learned has a dedicated region separate from Latin America. Riot has not mentioned any specifics on how it plans to step up moderation for what is likely to be a hotly used feature. For more [music] discussion on the tech news of the day, subscribe to the Daily Tech News Show at dailytechnewsshow.com, where you can also find the show notes and links to [music] every headline. Please remember to rate and review Daily Tech Headlines on your podcast service [music] of choice. From everyone here at Daily Tech Headlines, thanks for listening. >> [music] >> Mhm.