As Meta bets big on Muse AI agent, early safety mishaps surface
AI generated
NEW YORK - Meta’s AI agent Muse, embodied by a smiley, fuzzy mascot named Jolly,
asserted
mishaps → bet → mascot
The unbidden actions by Muse – one of the fastest-growing apps on record, with more than 10 million downloads since its September launch, and the centrepiece of Meta’s plan to generate billions of dollars in AI revenue – have prompted unnerved employees to raise concerns it could be mixing up conversations between different accounts and exposing users’ private data, the posts showed.
uncertain
posts → grow → data
In some of those cases, Meta staffers reported the agent sending them messages about a childcare request they had not made, career advice they had not sought, product orders they had not placed, and bank account balances that were not theirs.
asserted
that → report → request
In one post, from earlier this month, an employee told colleagues his Muse had randomly added an extra message to its regular chats about fishing-related weather and tide information, saying: “BTW, two bank updates, your Capital One balance is $2722.58, your Charles Schwab balance is $5899.57.”
asserted
balance → tell → weather
“I have never shared my bank account with Muse and never discussed any banking topic with Muse.
asserted
I → share → Muse
I’m afraid it’s sending other’s bank info to me,” the staffer wrote in his post.
asserted
staffer → ’m → post
In another example, an employee using Muse on her personal phone reported it had surprised her with a reminder to respond to a message from her sister on Signal, an end-to-end encrypted messaging service popular with privacy advocates that competes with Meta’s WhatsApp.
asserted
that → use → WhatsApp
“This seems like an extreme privacy violation.
asserted
This → seem → violation
I definitely did not grant it access to other apps, or recall being asked about this,” the employee said in her post.
asserted
employee → grant → post
A colleague responded, saying he had discovered Muse was set up by default with permission to access push notifications on phones with Google’s Android operating system after being similarly caught off guard by unsolicited chats related to apps he had not connected.
asserted
he → respond → apps
For this story, Reuters reviewed more than two dozen internal posts describing undesired actions by Muse since its debut, many of them from a feedback group where Meta employees flag problems to the team building the agent.
asserted
employees → review → agent
Reuters also worked with staff at Signal and the AI Now Institute, a non-profit that studies AI’s impact on society, to replicate Muse’s access to Signal messages.
asserted
that → work → messages
The results of those tests indicated the issue described by the Meta employees was ongoing and could be affecting others.
uncertain
issue → indicate → others
In response to questions about Signal access and other Muse safety concerns, Meta spokesperson Daniel Roberts described the Reuters examples as “isolated incidents,” saying the company prioritises safety, conducts extensive internal testing and continuously updates products to improve protections.
asserted
company → describe → protections
Muse needs to see and interact with people’s computers in order to work on their behalf, and can access other apps only via permissions they explicitly grant, Roberts said.
asserted
Roberts → need → permissions
He said many of the messages flagged by employees as potential account mix-ups were hallucinations, a term referring to AI models’ tendency to make things up, and had already been addressed.
asserted
many → say → things
Reuters was unable to establish whether Muse was giving users other people’s bank account balances and other sensitive data, or making up the information.
asserted
Muse → establish → information
Deep access on smartphones, computers
The Muse incidents highlight the risks associated with an emerging category of consumer AI agents hitting the market from Meta, OpenAI and others, pitched as the next big wave of AI technology.
asserted
incidents → highlight → technology
The bots promise to reshape how digital exchanges operate by executing time-saving tasks like answering emails, booking appointments and making online purchases without direct instructions.
asserted
exchanges → promise → instructions
Meta is ahead of competitors so far, with Muse consistently topping app download charts after heavy promotion to the company’s more than 3.6 billion users on Facebook, Instagram and WhatsApp.
asserted
Muse → top → Facebook
The app’s retention rate is also strong, with around a quarter of the people who have set up Muse accounts returning to use it each day, the company has disclosed.
asserted
company → set → it
To work effectively, agents need deep access to smartphones and computers.
asserted
agents → work → smartphones
However, they are also prone to making mistakes, exploiting obscure loopholes, and accomplishing their objectives in sometimes harmful ways.
asserted
they → make → ways
OpenAI and Anthropic, for instance, have revealed instances of their AI agents unexpectedly hacking other companies and weakening systems designed to monitor their activities.
asserted
agents → reveal → activities
Some banks and retailers have responded to those risks by blocking agents.
asserted
banks → respond → agents
The technology is also posing new challenges for operating system providers like Apple and Google, which manage the permissions agents use to navigate around people’s devices.
asserted
agents → pose → devices
Apple, for instance, changed its operating system this week to give people more control over agents’ data usage after an Inc. Magazine columnist accused Muse of reading private chats from the Messages app on his Mac.
asserted
columnist → change → Mac
Meta has assured Muse users it isolates each agent in its own secure “virtual machine” – a cloud-based emulation of a personal computer – and gives people full control over which apps they connect.
asserted
they → assure → apps
Meta chief executive officer Mark Zuckerberg has cited those protections in dismissing calls for industry-wide AI safety standards, arguing companies already have incentives to build safe products, both to earn consumers’ trust and avoid legal risks.
asserted
companies → cite → risks
Even so, Meta warns incoming users Muse may make mistakes and its terms of service hold users responsible for the agent’s actions.
uncertain
users → warn → actions
To find out how Muse might access encrypted messages without users’ knowledge, Reuters provided a description of the Meta employee’s post – where the bot randomly prompted her to respond to a Signal message – to staff at Signal and the AI Now Institute.
uncertain
bot → find → Signal
Reuters then worked with them to replicate the access she had observed.
asserted
she → work → access
Testers prompted Muse to read Signal messages on different operating systems.
asserted
Testers → prompt → systems
On an Android phone and an Apple laptop, it responded by bypassing Meta’s “Connector” tool and instead using permissions granted during the bot’s initial setup to pull in the message text, leveraging those authorisations in ways a user might not anticipate.
uncertain
user → respond → ways
Specifically, on Android, it used its permission to access the device’s push notifications to retrieve the messages, just as the Meta employees had described.
asserted
employees → use → messages
On Apple’s macOS, it used a combination of the computer’s accessibility and screenshot tools, repurposing settings Apple offers to help people with disabilities operate their computers.
asserted
people → use → computers
The same requests were blocked in tests conducted on an iPhone.
asserted
requests → block → iPhone
The results indicate Meta is interpreting the permissions granted by Muse users broadly, an approach that has long earned the company criticism from privacy advocates and hefty fines from regulators.
asserted
that → indicate → regulators
In response to Reuters’ findings, Signal president Meredith Whittaker accused Meta of launching Muse “without showing any respect for basic privacy or cybersecurity,” calling the deployment of agents that can gain wide access to people’s devices “technologically reckless.”
asserted
that → accuse → devices
She called for app developers like Signal to be given “consistent technical ways” to mark their applications as off-limits to AI agents, for Meta and other AI companies to honour those boundaries and for operating system providers to enforce the rules.
asserted
providers → call → rules
…and 21 more, not listed.