Experts at Gartner say Muse AI’s successful deployment would require strong governance around consent, data handling and accountability
Even as Meta’s new personal intelligence tool Muse takes AI to the next level -from answering queries to completing tasks- experts say the technology’s biggest test will be its ability to acting on the users’ behalf without compromising their data or triggering unauthorised transactions.
The potential of Muse is closely linked to the safeguards governing its autonomy and its roll-out in India will require clear governance framework around consent, data handling, and also compliance with DPDP and RBI requirements, making it a long drawn-out process, analysts add.
Meta Muse is a personal AI agent designed to help users complete tasks across the web. Among other features, Meta says it can interact with websites through a private browser, use stored credentials, access information such as email, and carry out actions including online purchases. The company is rolling it out only in the US market currently.
Niyati Daftary, Principal Analyst, Gartner, says Meta has put in additional protection within Muse, successful deployment would require strong governance around consent, data handling and accountability. “It is also important to have clear user awareness of what data the agent can access, how it is used and what actions it is authorized to perform, while remaining compliant with India’s DPDP and RBI requirements.”
Meta says Muse operates within an isolated virtual machine, with encrypted credentials and a Sentinel system that helps prevent sensitive information from leaving the protected environment. Dr Srinivas Padmanabhuni, CTO of AiEnsured says India was left out of the first rollout for a reason.
“Draft rules want visible human oversight and data that stays put. Until Muse can pass those tests in public, not just in a lab, the country may sit this wave out and watch the new transactional layer of the internet get built elsewhere,” he said. He lists agent stalling, picking the wrong tool or firing a payment nobody authorised among key risks with personal intelligent tools.
However, for a country like India where smartphone use is rising, benefits from Muse could be many.
Muse AI is a welcome step because users actually need AI to really work for them, Pawan Prabhat, Co-Founder of Shorthills AI, said.
Given that half of the world’s population use Meta’s apps in one form or the other, if they’re able to get it right, it might really be one of the most used personal AI in the world, he added. “There are questions about privacy and data leaks, but a lot of guardrails seem to have been put in to safeguard individual data.”
Vivekanand Gopalan, Chief Product Officer, Indusface, lists out specific guardrails needed before this goes live in India.
These include per-app permissions that expire rather than one blanket grant, mandatory human confirmation for any payment or irreversible action, and a plain-language audit log so users can see what the agent did and why. “Data residency and consent aligned to the DPDP Act are also essential,” he adds.