Meta's AI can now unfollow Instagram users who don't follow you back
A built-in Meta assistant turns Instagram's follow-back etiquette into an automated purge, and the users testing it are revealing how AI will handle the awkward parts of social life.
Muse, Meta's AI assistant, can track and unfollow Instagram accounts that do not follow a user back, and WIRED reports some people are taking full advantage. For decision-makers, the feature is an early test of how AI agents will automate socially sensitive choices, with consequences for platform trust, user consent, and product risk.
Muse, Meta's AI assistant, can track Instagram accounts that do not follow you back and unfollow them, WIRED reports. That is the whole trick, and it is less trivial than it sounds. Instagram's social graph has long run on reciprocity: a follow is a small signal of interest, and a follow-back is the polite reply. The new feature turns that etiquette into an automated action. Instead of scrolling through a follower list and making judgment calls one by one, a user can delegate the pruning to an AI assistant. According to WIRED, some people are taking full advantage.
The hook is not that Instagram added another settings toggle. It is that an AI assistant is now making a judgment about a relationship and executing it. Muse is not merely suggesting a caption or summarizing a message; it is acting on the social graph. Unfollowing is a small action with real social meaning. It can signal disinterest, clean up a ratio, or settle a score. When an assistant handles that choice, it moves AI from content generation into social maintenance, a far messier territory.
For creators, founders, and anyone who treats Instagram as a professional surface, follower counts and follow-back ratios function as social proof. A clean ratio can look like influence; a messy one can look like neediness or neglect. The appeal of an automated unfollow tool is obvious: it promises a tidier profile without the awkward manual work. The risk is equally obvious. If everyone can automate the pruning, pettiness becomes scalable. A feature that helps one user tidy up can help another user conduct a quiet, systematic purge of people who did not reciprocate.
Meta's incentive is to make its AI assistant feel useful, not just conversational. A helper that can manage a calendar, draft a message, or clean up a follower list is more likely to become a daily habit. But every automated action on a social platform creates a governance question. Did the user intend to unfollow that person, or did the assistant interpret a stale follow as a slight? Can the action be undone? Is the other person told? Instagram has long offered controls for blocking, restricting, and privacy, but delegation adds a new layer: the user is not just changing a setting, but authorizing an agent to act on their behalf.
That is where the feature becomes a test case for AI agents more broadly. In productivity software, agents are already being pitched as tools that can draft emails, schedule meetings, and complete multi-step tasks. On social platforms, the tasks are more emotional. An AI that books a meeting is helpful; an AI that decides which relationships to cut is making a judgment with reputational consequences. The platform has to decide how much agency to hand over, how transparent to make the action, and how to prevent the assistant from becoming a weapon for petty behavior.
For Meta, the product calculus cuts both ways. The feature makes Muse more capable and gives users a reason to open the assistant inside Instagram. It also creates a new surface for backlash. Social platforms have spent years learning that rejection features can escalate quickly, because they turn private feelings into visible actions. An automated unfollow is not a public declaration, but it can still be felt. If users believe the assistant is encouraging score-settling, the feature could become a symbol of AI-enabled pettiness rather than a smart convenience.
The regulatory backdrop is still forming. AI assistants that act on a user's behalf sit at the intersection of privacy, consumer protection, and platform competition. Regulators have been increasingly focused on automated decision-making and the ways digital tools can nudge behavior without users fully understanding the trade-offs. This particular feature is small, but it fits a larger pattern: AI is moving from answering questions to taking actions. The companies that get delegation right will be the ones that make the agent's behavior legible, reversible, and clearly bounded.
For executives beyond Meta, the lesson is to watch the mundane use cases, not just the headline-grabbing ones. The Instagram feature is a small case with a large signal: AI is entering the messy, emotional layer of relationship management. Founders and product leaders should ask whether their own AI features are automating tasks that users would want to review, undo, or explain. The competitive advantage may not go to the assistant that does the most. It may go to the one that knows when to pause and ask. On Instagram, the question is whether users want a valet for their social graph or a referee.
This story's Key Insights and Take-aways are locked.
Create a free account to unlock Executive Actions for one credit.
Register to UnlockAlways free for Executives Club members. Join the Club
More in Technology
Anthropic mathematician claims Claude helped crack the Riemann hypothesis
If the proof holds, AI has toppled math's most famous open problem; if it does not, it is a case study in announcing breakthroughs before anyone verifies them.
AI's $4 Trillion Wipeout Erases Nasdaq's Gains for the Year
A brutal week took roughly $4 trillion off the AI sector and left the Nasdaq flat for the year, and the damage ran the full length of the supply chain.
Anthropic mathematician says AI proved the Riemann hypothesis. Verification will take time.
A single unverified claim from inside Anthropic now sits against more than 160 years of open mathematics, and the gap between asserting a proof and surviving peer review is the entire story.


