When Backlash Is Actually a Boundary

Welcome to Briefly AI, a podcast by Harry Sharman, created by AI and voiced by an AI synthesis of Harry Sharman. So if this sounds like Harry, that's the point. If it sounds clever, blame the machine.

This week, Meta launched an Instagram feature that let users generate AI images using content from public accounts — without asking the account owners first. Within 48 hours, they pulled it. The backlash was fast, loud, and effective.

Now, on the surface, that looks like a straightforward story about consent. And it is. But here's the more interesting question underneath it: why did that land so badly, when people have quietly accepted dozens of AI features that are, in some ways, just as intrusive?

Because the answer isn't really about privacy law or data policy. It's about something older and more instinctive than that. It's about identity.

Right, let's get into it.

The Meta feature — they called it Muse — worked like this. You tagged a public account in a prompt, and the system would generate new AI images inspired by that person's visual content. Style, aesthetic, look. If you had a public account, your creative identity was, by default, raw material for someone else's imagination. Opt-out, not opt-in.

The backlash came immediately from creators, photographers, artists, small businesses. Not primarily from privacy advocates. Not from regulators. From people whose professional identity is literally embodied in their visual output. Their response wasn't "this is a data breach." It was something closer to: "that's mine."

And within two days, Meta folded.

Now compare that to what most people accept without complaint. Their email habits being analysed to suggest replies. Their photos being scanned to auto-sort albums. Their search patterns feeding personalisation engines that know their interests better than their colleagues do. All of that happens quietly, in the background, and most of us have made peace with it.

So what's different here?

Here's the behavioural science reading. There's a distinction that psychologists draw between what you do and who you are. Your search history is what you do. Your creative output — especially if it's how you earn a living or how you present yourself to the world — is much closer to who you are. The moment AI starts operating on that territory, you're not just dealing with data any more. You're dealing with identity.

Harry Sharman wrote about this in April, framing it around workplace adoption specifically: the biggest barrier to AI at work isn't skill, it isn't knowledge, it's what the tool appears to threaten in people's sense of professional self. The Meta Muse reaction is that thesis playing out in public, at speed, in front of everyone.

And there's something else worth naming here. The people who pushed back loudest weren't the people least comfortable with AI. Many of them use AI tools constantly — for editing, for ideation, for production. They're not anti-AI. They're anti this particular use. Which tells you the resistance wasn't about the technology. It was about the boundary. Where does my contribution end and your raw material begin?

That's a genuinely hard question. And right now, there's no shared answer to it.

Which brings us to the second thread this week. Because running alongside the Muse story is research that keeps finding the same uncomfortable pattern in workplaces: roughly half of workers — studies put the figure between 45 and 48 percent — are hiding their AI use from colleagues and managers. Not because they're doing something wrong. But because they've learned, usually by watching what happens to other people, that admitting to AI use attracts a social penalty. You get rated as lazier. Less capable. Less likely to be handed the interesting work.

So they do it anyway. They just don't say so.

The connection to the Muse backlash isn't obvious at first, but stay with me. In both cases, what you're watching is people drawing a line around the cognitive work they want to be seen to own. In one case, it's creative output on social media. In the other, it's professional output at work. The specific context is different. The psychology is the same.

We have a very deep, very human need to be legible to other people as the person who did the thing. That need doesn't disappear because AI gets more capable. In some ways, the more capable AI gets, the more urgent the need becomes — because the boundary between "I made this" and "I fed a prompt" gets thinner.

And here's what makes this hard for organisations — and for individuals — to navigate. The research on disclosure stigma suggests that the penalty for admitting AI use is, in many workplaces, genuinely real. It's not paranoia. People are responding rationally to a social environment that hasn't caught up with actual practice. Which means you end up in this strange situation where AI adoption is widespread and largely hidden, managers are making decisions based on a version of reality that doesn't exist, and the gap between what people say they do and what they actually do keeps widening.

That's not a cultural quirk. That's a structural problem.

The Meta story gives us a slightly more hopeful signal, actually. Because what the Muse backlash showed is that when the boundary violation is visible and immediate — when people can see exactly what's being taken and by whom — the response is fast and it works. Meta didn't need a regulator. They needed enough creators to say loudly, in public, "not that." And the feature was gone in two days.

The workplace version of that is harder, because the boundary violations are subtler and the power dynamics are less symmetrical. You can't cancel your employer in 48 hours. But the underlying principle might be the same: the clearer people can be about which cognitive work they want to remain legibly theirs, the better chance they have of protecting it.

So the question worth sitting with this weekend is this. When you use AI — at work, in your creative practice, anywhere — what are you comfortable having attributed to you, and what would feel like a misrepresentation? Not legally. Personally. Because most of the conflict we're seeing right now, the backlash, the hiding, the resistance, isn't really about capability or cost or even trust in the technology. It's about where people draw the line between the tool and themselves.

And most of us haven't answered that question clearly yet. Probably worth doing, before someone else answers it for you.

And that's Briefly AI. An AI made this. Harry Sharman thought it was a good idea. You've just decided whether they were both right. Back tomorrow — subscribe so you don't miss it.