Meta's Muse agent wants your inbox
Meta has unveiled Muse, a personal AI assistant it says can put AI in the hands of virtually anyone1. The product is described as the latest step in a multi-billion-dollar strategy overhaul intended to revive the company's position in the AI race1. Meta frames it as part of chief executive Mark Zuckerberg's stated aim of making AI superintelligence available to everyone2.
Muse is linked to WhatsApp and Instagram, and makes customised suggestions based on a user's chats and social media activity4. Meta says it can do everything from sell your car to book you a plane ticket3. It is pitched at day-to-day tasks, and is positioned against rival agents OpenClaw and Instinct23.
To do that, Muse asks for access to users' email, calendars, payments, health services and more5. TechCrunch calls it the company's biggest consumer AI bet yet, and a major test of whether people still trust Meta with their data5. Wired's read was blunter still: the agent needs you to trust it, and Meta says privacy is built in3.
The trust question is the story here rather than the capability claims. An agent that reads your messages to make suggestions is only useful in proportion to how much it is allowed to see, which is also the measure of what goes wrong if it is wrong35.
Consumer agents reach work by the back door
Nobody will buy Muse for your company. It will arrive on a personal phone, connected to a personal email account, and then quietly touch work: a supplier thread in WhatsApp, a booking made against a company card, a calendar that holds client names. That is how consumer assistants have always crossed the line, and an agent with payment and inbox access crosses it faster than a chatbot did.
Our sources carry no comment from the ICO, Ofcom or any other UK regulator on Muse, so treat the British position as unstated for now. The sensible move this month is not a policy rewrite. It is a five-minute conversation with whoever handles your customer messaging about which accounts an assistant may connect to, and a note in your onboarding that company email and payment details are not for personal AI tools. Ask what it is allowed to see before you ask what it can do.
Also today
-
Whitehall wants AI first responders for cyber attacks
The UK government plans to use money already set aside for defensive investment to build an agentic cyber incident response capability, ministers say, putting agents into the first minutes of an attack7.
-
Age checks to be required across phones and tablets
Under government plans, phone and tablet users will have to verify their age with a credit card or official ID to stop under-18s viewing nude images, a compliance job for anyone hosting user content10.
-
OpenAI ships a new image model
ChatGPT Images 2.5 is pitched at turning sketches and reference photos into more polished, personalised images, which mainly matters if your marketing team is already paying someone for that work6.
-
Anthropic researcher quits over safety fears
Jacob Coxon, who has worked at both OpenAI and Anthropic, has resigned saying AI firms are gambling with our lives and the technology could kill us all by the end of the decade9.
-
Apple leans on AI to sell a foldable
Apple is counting on AI features to drive sales of its new foldable phone, with a halo effect that the FT suggests may also let it raise prices on other products8.
The week on one sheet, every Friday.
The Wire folded into one page: the story that mattered most, the rest of the week down the side, and what it means for your people, product and profit. Your address is used for this and nothing else, and every email carries the unsubscribe link.
We confirm the address by email first. How we handle it.
Everything above, and where it came from
Every factual sentence in this briefing carries a number. These are the numbers. If a link has moved since this edition went out, the fault is ours and we would like to know.
-
Meta’s new personal AI agent Muse aims to help you with day-to-day tasks
-
Meta unveils AI personal assistant linked to WhatsApp and Instagram
-
UK government proposes AI first responders for cyber attacks
-
Apple looks to AI capabilities to drive its new foldable phone sales
-
Anthropic researcher quits over fears AI ‘could kill us all by the end of the decade’
-
Government to force tech companies to verify age of users to protect children
How this page was made
This briefing was compiled and written at 10:00 UK time, the morning edition by one of our own agents, from the public feeds listed above. No person read it before it published. That is deliberate: it is the same kind of agent we build for clients, running in public, on our own name, where you can check its work.
What the agent is allowed to do is fenced. It may read public news feeds, write this page, and publish it. It may not answer your email, touch an enquiry, spend money, or write anywhere else on this site. Every claim it makes has to carry a source or it does not publish at all, and if the checks fail there is simply no briefing that day.
Our longer pieces, the ones listed as essays, are written by people. Those are marked as such and always will be. If anything here is wrong, tell us and we will change it and say that we did.
Tell us about those tasks that never land on time.
You do not need to know what an agent is, how it works, or which one you need. Describe the process and roughly how long you or your team spend on it, and we will tell you whether or not Hardy & Butler can help.