Meta settled for $17.1 billion. The historians say the tobacco moment never happened. I’ve seen this from the other side of the table.
Brussels, September 9th 2026.
In 2010 I was a strategist in a Paris consultancy that did, among other things, what is politely called public affairs. One of its clients made the world’s best-known soft drink. Across Europe, governments were talking about taxing sugar, and the client’s view was that the conversation could be managed: some lobbying at the highest level, some visible commitment to health and the environment, and the tax would not come.
It came. France adopted its soda tax in the 2012 budget. Hungary had moved a year earlier, Belgium and the UK would follow. Nothing substantial changed in the product.
What we proposed, when the lobbying stopped working, was not a solution and we knew it. We proposed to monitor public opinion for them: track the trends, read the signals, help them anticipate the next shift before it became a tax. It sounded like foresight. It was a retainer. It was the answer a consultancy gives when the true answer, that the product is the problem, is not one the client is paying to hear, and not one that renews.
And the thing I remember most is not the tax, nor the retainer. It is that in two years of meetings I never met anyone inside the company who did not believe, sincerely, that the product was hydration. Not a line they had been given. A belief. The vocabulary had arrived long before the tax, and it had arrived from inside. We were not there to challenge it. We were there to help them hear it coming.
I thought of that in late August, reading the coverage of Meta’s settlement with 47 American states and the District of Columbia: up to $17.1 billion over a decade, and changes to Instagram and Facebook to make them less compulsive for the young. Texas settled separately the same day, for another billion. The states had accused the company of marketing an addictive product to children. Meta itself had put the number $1.4 trillion in a filing, roughly its own market value, to say what the exposure could be. Then it settled for a little more than one per cent of that, and its shares rose one per cent on the news, which on a company that size is about what the settlement cost. The market, in other words, paid the fine.
The critics wanted a tobacco moment. The historians the New York Times went to see told them there had never been one. The 1998 Master Settlement, $206 billion over twenty-five years, was in the tobacco companies’ interest: they bought their way out of the litigation and kept selling. What changed smoking was slower and less cinematic. Twenty-five years of activists, municipalities, employers and eventually parents making the product socially unacceptable, one workplace and one restaurant at a time. The settlement arrived after the culture had already turned, and took the credit. Scott Galloway would add, and he is right, that the culture did not turn on its own: cigarette taxes rose fourfold at the federal level and sixfold in the states, and the product was age-gated like alcohol and guns. The incentives moved first. The culture followed them, and the vocabulary followed the culture.
If that reading is right, and I think it is, then the question about social media is not whether the fine was big enough. It is where we are on the slow curve. And here is what my two years with sugar taught me: the curve has a stage that nobody in the courtroom sees, which is the stage where the category rewrites its own vocabulary. Not for the regulators. For itself. “Hydration” was not a lie told to us; it was a thing the company had come to believe, and the belief is what made the product defensible to the people who sold it. Wellbeing. Connection. Time well spent. Teen accounts. Read them again after the settlement and hear what they are: the same move, made by a category that cannot change its product because the product is working exactly as designed. Jonathan Haidt pointed out, the week of the settlement, that the recommendation engine itself is untouched: still built to maximise young people’s engagement, whatever the content does to them. Sugar tax, same product. Settlement, same algorithm.
And the vocabulary did not wait. The day after the settlement, Meta launched an advertising campaign calling on YouTube and TikTok to adopt the same protections, and made five billion dollars of its own penalty conditional on their doing so. Galloway, who teaches brand strategy, calls this laddering: highlight your strength, de-position the competitor. The company accused of hooking eleven-year-olds spent the following week positioning itself as the industry’s leader on child safety, and it did so with the tools of my trade. Somebody wrote that campaign. Somebody was proud of it.
One of the documents that surfaced in the Meta cases was an internal message about targeting eleven-year-olds. “We have to hook them young,” the employee wrote, adding that it felt like the tobacco companies of a couple of decades ago. We are, as Antonio Nieto-Rodriguez wrote the week of the settlement, writing right now the documents that someone will one day read aloud in a courtroom.
I would go one step further, because I have been on the payroll. It is not only the internal documents. It is the external ones, and the advice. The positioning lines, the manifestos, the “we are a health and wellness company now”, and the monitoring dashboards that let a company watch the culture turn against it in real time while changing nothing. Someone writes those, and someone sells them. In 2010 it was people like me.
A company whose product works as designed has three moves. It can change the product, and accept a smaller, less compulsive business; no one at that valuation will. It can change category and say so, “we were an attention company and we are leaving”; no one says that either. Or it can change the label and keep the engine. That is the move tobacco made with e-cigarettes, and it is the move Meta is making in front of us: it no longer calls itself a social media company. It is an AI company. The uncomfortable part is that this is not only vocabulary. Meta has built something: the infrastructure, the graph of three billion people, the capital to spend a settlement on Tuesday and a data centre on Wednesday. The pivot is real. What travels with it, unexamined, is the design priority that got them to Oakland. Same incentives, new product, a new decade before someone reads the documents aloud. The diagnosis is not hard. What is hard is being the one in the room who says it.
So the question I want to leave with you is a practitioner’s question, and it comes in two directions.
If you run a company whose brand lives mostly on these platforms, on rented land that is now, slowly, being reclassified as a public health matter: what have you built that would still stand if the ads carried warnings, or stopped? Brands that lived on cigarette advertising had to answer this in a hurry. The ones that survived had something underneath the media.
And if you are the one being asked to write the new vocabulary, the consultant, the agency, the strategist: which side of the courtroom are you writing for? I did not ask myself that in 2010. I left in 2015, and not for that reason, or not only: a child, a ceiling, the wish to build something of my own. The years since have been wild, and bittersweet more than once. But of everything they’ve given me, the thing I hold onto most is this: I have not written a sentence I did not believe, and I have not worked for a company I consider harmful. That was not a moral achievement. It turned out to be a structural choice.
Beatriz Vílchez Silva · La Colmena Design, Brussels lacolmenadesign.com