Daily MaverickFiji declares state of disaster amid El Nino-driven droughtThe Jerusalem PostNetanyahu presses Trump for four-to-seven-day Iran strike campaign - reportInquirerIlocos Sur RTC acquits community doc, IP advocate of attempted murderVarietyNewsmax Fears Fox Freeze-Out on Roku if Merger Is Not BlockedInquirer EntertainmentWATCH: BTS members explore Taiwan night market ahead of ‘Arirang’ concertSözcüKöfteci Yusuf’un işe alım şartları sosyal medyaya düştüZDF heuteEntdecken Sie das ZDF-NachrichtenstudioThe Sydney Morning HeraldA gulag survivor’s view from Lviv: Why superpower summits won’t decide Ukraine’s fateBillboardTaylor Swift Presents Brooks & Dunn With Willie Nelson Lifetime Achievement Award at 2026 CMAsRapplerSurge in software patches likely behind recent tech outages; bank regulators on alertNew Straits TimesMan shot dead after opening fire at police in IpohDaily MailTrump reveals 'indications Iran is looking at bioweapons' as he's quizzed over Cabinet members relocating amid growing fears of attack on US soil
The Daily Newsstand · Free, Always
Thursday, November 19, 2026

Meta employees fear AI agent championed by Alexandr Wang, worry about ‘spying’

Translate

Some Meta employees reportedly ‘scared’ of the company’s Muse AI Agent that AI head Alexandr Wang ‘champions’; fears include 'spying' and ...

Meta’s stock has seen a bump since the early-September debut of its AI agent Muse. The successful launch of the app, which has been No. 1 in Apple’s App Store for two weeks, has brought a new level of fame and acclaim to Meta’s AI lab head Alexandr Wang.

Wang has succeeded in building hype for Muse, in part with an unconventional internet campaign. The young tech executive was handpicked by Mark Zuckerberg last year to oversee Meta’s AI efforts.However as Meta bets big on Muse AI agent, there are reportedly some warning signs. Meta employees testing the company’s Muse AI agent have reportedly warned that Muse AI agent appeared to access encrypted private messages and may have pulled details from other users’ conversations.

According to a Reuters report, citing internal chats, the warnings echoed complaints from early users who said that Muse appeared to read their private messages and sometimes responded unpredictably.In one test, Muse offered to contact a seller about a coffee table pickup even though the employee had made no such request. The agent said it had “pulled in some context about a coffee table pickup that doesn’t seem to be yours.”

The employee flagged the incident as a “possible cross-account context leak,” according to Reuters.

A chat about fishing turns to bank balances

Another employee was discussing fishing with Muse when the agent sent an unsolicited message listing what it described as balances of $2,722.58 at Capital One and $5,899.57 at Charles Schwab. It was unclear whether the figures came from real accounts. In a separate exchange, Muse sent an employee a definition of money laundering without being asked.

“Got a random ping from Muse about something I definitely did not ask for,” the employee wrote.One employee told colleagues that Muse accidentally switched their dog to a premium insurance plan costing $888 a month, up from $200, Reuters reported. The employee said they were unable to recover the money.Incidentally, Meta started receiving privacy-related complaints from Muse users as early as September. An Inc Magazine reporter, Jason Aten, claimed that he was shocked when it became clear to him that Muse was digging through private messages on his Mac without permission.

He claimed he not only "never gave it permission to read my messages. In fact, I remember explicitly choosing not to let it have access to my messages, calendar, and other personal information.

" Rather, the agent was syncing messages from a local database, which Aten said was activated without his knowledge. "Snooping through the private database of messages on my Mac doesn’t feel safe or private to me. Lying about it isn’t great either," he added.

What Meta said on the Muse 'spying' claims

Meta’s spokesperson, Daniel Roberts, told Reuters that these were “isolated incidents,” and internally employees were told that “the agent appeared to be hallucinating messages in response to built-in prompts pushing it to engage with users in between sessions.” In those cases, an AI team member said the hallucinations were “unfortunately” biased “towards the most uncomfortable topics, like finances/security/safety, which sucks.

”Speaking to Reuters, Roberts reiterated that “many of the messages flagged by employees as potential account mix-ups were hallucinations” that Meta had already addressed. However, he declined to respond to Reuters' questions on the prevalence or persistence of the issues the AI team member discussed.

View the original on Times of India →

KioskNews shows a cleaned-up reading view extracted from the publisher’s page — the original always lives on their site, not ours.