Meta’s Muse can take on tasks such as buying groceries and negotiating bills, but a BBC report argues that AI agents doing this at scale could strain the websites and services they use. The question is no longer only what an agent can do for its owner, but what happens when millions of them arrive at once.
Meta AI Watch analysis
What happened
The BBC’s Thomas Germain tested Muse on tasks including contacting a Marketplace seller, ordering dental floss and negotiating a software-subscription discount. The report says Muse can use a browser to carry out tasks, and that Meta says it asks permission before purchases and is designed to respect websites’ interests.
Researchers interviewed by the BBC warn that agents could swamp businesses with enquiries, compete for scarce tickets and appointments, and put pressure on websites built for people rather than tireless automated visitors. The report cites an analysis finding web traffic from bots rose 124% in the year to June 2026. It also describes a security flaw that researcher Patrick Wardle said could have let an attacker hijack Muse and access connected devices; Meta says it fixed the flaw and that the app underwent further security testing before release. Read the BBC report.
Why it matters
An agent that saves one person time can create extra work for everyone on the other end. Businesses may face a flood of automated requests, while people could find tickets or appointments harder to get if bots can act faster and in greater numbers. The BBC’s report describes risks and expert forecasts, not proof that every predicted disruption is already happening.
Our read
The useful test for agents is not just whether they complete a task. It is whether they do so without shifting the cost onto websites, businesses or other people. Giving an agent permission to act is one side of the bargain; clear limits on how it behaves online are the other. The internet may need rules for agents before it becomes an obstacle course of bot checks and queue-jumping.
What to watch
- Whether websites introduce new limits or fees for automated agent activity.
- Whether regulators or platforms set clear rules for agents making bookings and purchases.
- How Meta explains Muse’s permissions and the security flaw described in the BBC report.
Discussion spark: If an AI agent gets you a ticket or appointment faster, should the service be allowed to limit or charge for its automated requests to protect access for everyone else?
Sources and evidence
- 'Things may get ugly': Meta's new AI Muse is about to make the internet more annoying – BBC (1 October 2026, 09:08 UTC)
not affiliated with, endorsed by, or operated by Meta