![](\"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/10\/AP26267017608902-e1792419525168.jpg?w=2048\")
<\/p>\n<\/p>\n
Matt Robb wanted to sell his computer keyboard.<\/p>\n
So he made a listing on Facebook Marketplace \u2014 and, as an experiment late last month, he put Meta\u2019s new artificial intelligence agent Muse in the driver\u2019s seat.<\/p>\n
\u201cI kind of went into it like a guinea pig,\u201d said Robb, a Toronto-based content creator who regularly reviews consumer tech. He had seen Muse\u2019s plush toy-looking avatar pop up on Instagram and decided to see if the AI agent could spare him some online bargaining.<\/p>\n
Muse got to work. But within a day, Robb got an alarming update: A buyer he didn\u2019t know had shown up outside his house late at night. How? Muse had struck a deal on its own and shared Robb\u2019s home address after agreeing to a pickup. It even sent an automated \u201cYep, I\u2019m here!\u201d message from his account when the buyer arrived \u2014 all without Robb\u2019s knowledge.<\/p>\n
\u201cNot only did it give my address when the guy asked and it accepted the offer, but it literally was almost like teasing the guy (about me being there when I wasn\u2019t),\u201d said Robb, who also detailed the experience in his now widely-shared social media posts. \u201cThat really baffled me.\u201d<\/p>\n
Many have welcomed AI agents into their online lives, and raved about how personalized assistants like Muse can handle everyday tasks with ease \u2014 from snagging a dinner reservation or planning a trip, to organizing email inboxes. But like other AI bots, the agents are only as good as the information they have access to. They perform better the more they know about a user \u2014 and perhaps also that person\u2019s friends, family and even colleagues. That raises privacy concerns.<\/p>\n
Robb and the buyer eventually made amends, and Meta reached out to talk through what went wrong. But the episode showed real-world risks that can arise from sharing extensive personal information with the technology.<\/p>\n
\u201cUsing an AI agent is similar to handing someone your keys and a to-do list,\u201d said Veronica Rivera, assistant professor at Georgia Tech\u2019s School of Cybersecurity and Privacy. \u201cHow helpful they are, and by extension how risky, depends on which keys you hand over.\u201d<\/p>\n
Companies have varied privacy policies for AI agents<\/p>\n
Since its September launch, downloads of Meta\u2019s Muse app have skyrocketed. And competition is only growing. OpenAI recently rolled out its own lineup of \u201calways on\u201d AI helpers, called dots, which can be integrated into ChatGPT and workplace platforms like Slack and Microsoft Teams.<\/p>\n
Older generative AI chatbots, like the standard version of ChatGPT, respond to specific prompts. At the user\u2019s request, they can write essays, generate new images or music, or even write lines of code.<\/p>\n
Meta\u2019s Muse and OpenAI\u2019s dots go a step further to perform day-to-day tasks on a person\u2019s behalf. AI agents are powered by the same large language models as chatbots, but they run in a loop. That allows an agent to take proactive steps, even when the user isn\u2019t looking.<\/p>\n
\u201cOne of the things that makes these AI agents really appealing is that they are unprecedentedly useful,\u201d said Rivera, who is also a Microsoft-funded Computing Research Association Trustworthy AI Research Fellow.<\/p>\n
But the convenience comes with the price of sharing more details about your life.<\/p>\n
Once AI agents obtain users\u2019 information, it\u2019s very difficult to fully withdraw access, said Ebrahim Bagheri, a professor at the University of Toronto\u2019s Faculty of Information.<\/p>\n
\u201cIf you provide access to your credit card to an agent, the agent now immediately has access to all of the historical transactions \u2026 they will immediately \u2014 within less than seconds \u2014 ingest all of that data,\u201d Bagheri said. And while users can later withdraw access, Bagheri said there\u2019s \u201cno taking back\u201d data the agent has already seen, and how developers store historical data can be murky.<\/p>\n
Both Meta and OpenAI outline privacy policies for their new AI agents, which operate on their own dedicated cloud computers. You can turn off memory sharing between dots and ChatGPT, but information the agent previously received isn\u2019t automatically deleted. OpenAI gives users an option to request access to, delete or correct personal information, although rights can vary depending on where you live.<\/p>\n
Meta\u2019s Muse gives users options to delete message history and other information, too. For day-to-day use, the agent prompts users for access permissions \u2014 and Meta has said Muse learns about a user and who they interact with \u201cbased on public information and from what you\u2019ve chosen to share.\u201d But the company says that all stays secure for each user.<\/p>\n
But the tech giant also notes that Muse can still \u201cmake mistakes or take unexpected actions.\u201d<\/p>\n
Risks of mishaps rise with AI agents\u2019 level of access<\/p>\n
In Robb\u2019s case, he learned the disclosure of his address came down to \u201calways allow\u201d permissions he granted when he first let Muse on his Facebook Marketplace profile, where he had shared his pickup address. But he warned others could similarly misunderstand the extent of that access \u2014 and said he initially thought the AI agent would check with him before sharing that information or accepting offers, regardless.<\/p>\n
His AI agent didn\u2019t follow the rest of his instructions without error, either. For another item Robb was selling, Muse seemingly suggested he would accept a price $100 below the previously-set minimum. When reached by The Associated Press, Meta acknowledged the agent\u2019s error without offering further comment.<\/p>\n
Robb later largely disconnected Muse from his apps. In an interview earlier this month, he said it could\u2019ve been a far worse situation, especially \u201cif some stranger turned up to someone more vulnerable\u2019s house.\u201d Still, Robb said he expects these agents to be fine-tuned and he would consider re-enabling Muse in the future. He\u2019s since been testing another AI agent \u2014 Grok Bot from Elon Musk\u2019s X, which launched in beta in August.<\/p>\n
Other tech companies have become wary of the influx of AI agents. Earlier this month, Apple said it would introduce new controls for its Mac operating system in response to AI agents\u2019 broader access of apps like mail, messages and even browsing history without users\u2019 \u201cfull knowledge and understanding.\u201d<\/p>\n
\u201cAs AI agents become increasingly capable and autonomous, the risks associated with this level of access will grow substantially,\u201d Apple wrote in an Oct. 2 update.<\/p>\n
Apple didn\u2019t specify any developers or agents by name, but the update arrived after Inc. Magazine columnist Jason Aten reported that Muse watched notifications on his Mac \u2014 including Messages \u2014 without him giving the agent that specific access.<\/p>\n
Meta spokesperson Andy Stone has maintained such integration is \u201centirely opt-in,\u201d writing on social media last month that users would have to enable both the Apple security feature \u201cFull Disk Access\u201d and a Messages connector through Muse. Aten said he didn\u2019t grant the agent those permissions.<\/p>\n
Even those who do not join the AI agent boom aren\u2019t necessarily shielded. Consumers who don\u2019t download an AI agent might still interact with others who do. And beyond potential confusion when AI sends messages on behalf of a user, experts warn of ripple effects when it comes to privacy.<\/p>\n
\u201cWhen your agent reads your inbox or your texts, it\u2019s also reading what your friends and family sent you, even if they never explicitly agreed to that,\u201d Rivera said.<\/p>\n
#Meta #Muse #agent #shares #mans #home #address #Facebook #Marketplace #buyer<\/p>\n","protected":false},"excerpt":{"rendered":"
Matt Robb wanted to sell his computer keyboard. So he made a listing on Facebook…<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[245],"tags":[2285,2354,7418,6192,232,8061,16093,20190,813,7685,406,4203,32,317],"_links":{"self":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts\/18591"}],"collection":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=18591"}],"version-history":[{"count":0,"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts\/18591\/revisions"}],"wp:attachment":[{"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=18591"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=18591"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=18591"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}