Why do AI companies want you scared of their own product?

📣😱 Dystopia as a marketing tactic? A little over a month ago, Diary of a CEO host Steven Bartlett invited sitting US Vice President JD Vance to discuss the Iran war, his upbringing, the inner workings of the White House, and AI — on the latter, Vance made an interesting comment (watch, runtime: 1:29:00 - 1:37:45). When asked about his worries regarding the AI takeover, the US vice president said: “For AI companies, there’s a certain incentive to be super dystopian because it’s a form of viral marketing. If people are scared of your product, that must mean it really works. If they’re not scared, maybe it doesn’t work all that well.”

Vance was talking about AI. But he had, perhaps without meaning to, described one of the oldest tools in the marketing book: the fear premium.

Feel something = buy something

Marketing has always traded in emotion. Joy, aspiration, belonging, nostalgia — the industry built itself on the science of making people feel their way to a purchase. The academic term is affective marketing: the deliberate use of emotional triggers to shape consumer behavior. Fear is one of the strongest triggers. The prospect theory — published in 1979 — dissects this best, demonstrating that people feel the pain of a potential loss roughly twice as intensely as they feel the pleasure of an equivalent gain. Losing EGP 500 stings harder than finding EGP 500 in the back of a drawer feels good, to give an example.

… Marketers have been exploiting this since.

The fear formula

Fear may sometimes backfire, however. In her 1992 paper, The Extended Parallel Process Model, communications scholar Kim Witte explains why fear — as a marketing tool — sometimes works, and sometimes fails catastrophically. Witte’s model identifies two responses a message built on fear can convey. The first is danger control: the person processes the threat rationally, accepts it as real, and takes the recommended action. This is the advertiser’s goal. The second is fear control: the person feels overwhelmed by the threat and, rather than acting, emotionally detaches — dismisses the message, ignores the risk, changes the channel. This is the advertiser’s nightmare.

The deciding factor between the two responses, Witte found, is not how frightening the message is. It’s whether the message pairs the fear with a clear, credible, achievable solution. Raise the fear without offering the exit, and you’re more likely to produce paralysis than purchase. The formula is: make them afraid, then hand them the key.

It’s how some of Egypt’s best advertising campaigns operated. In 2014, the National Bank of Egypt ran a brilliant campaign — brought to life by Tarek Nour — showing everyday Egyptians losing their cashmoney in a slew of domestic disasters, suitors being rejected for lack of funding, and other similar ludicrous — but not too much — scenarios that were plausible enough to incite fear without going overboard (watch, runtime: 0:44). The advertisements would remain a part of pop culture for years to come, with the phrase, “Why don’t you go check up on your money?” (Mat’oom tettamen ‘ala feloosak) etched into the Egyptian colloquial reference book.

And NBE offered a solution: Open an account. Prospect Theory and Parallel Process Model executed in under 45 seconds. One could even argue that the iconic “Never Say No to Panda” campaign, which became the first Arab commercial to gain widespread global virality, employed this methodology — though the threat of being hazed by a panda was far less realistic than losing your savings to a leaking pipe.

Industries run by fear

Fear marketing didn’t start with AI or Egyptian banks. Ins. companies have long operated on it. Cybersecurity firms, too. The commercials and campaigns for these industries have long capitalized on fear of destroyed cars, burnt down homes, or late-night burglaries. Threatened with jeopardy, consumers bought, bought, and bought. Healthcare is one of the more prominent industries employing fear marketing — exposed lungs and gangrene feet on cigarette packs being the clearer example (though smokers have since become desensitized).

Fear, dystopia, AI

Vance’s comment hits the nail right on the head: AI companies are benefiting from the dystopia narrative. In 2015, OpenAI CEO Sam Altman publicly said: “AI will probably end the world, but in the meantime, there will be great companies created with serious machine learning.” In 2019, OpenAI declared its GPT-2 model too dangerous to release — then released it nine months later, with Altman later acknowledging the alarm was overblown. In 2023, hundreds of tech leaders including Altman, Google DeepMind’s Demis Hassabis, and Bill Gates signed a statement declaring that mitigating AI’s risk of causing human extinction “should be a global priority alongside pandemics and nuclear war.” They made it. They fear it. They are still selling it.

The pattern is consistent across major AI labs: a company claims it has created an AI so powerful it is terrifying — too dangerous to release immediately, potentially catastrophic in the wrong hands. Then, after generating significant attention and positioning themselves as the responsible actors in a dangerous field, they release it.

This is where AI doom marketing departs from the NBE ad. Traditional fear marketing raises the threat, offers the product as the solution, and closes the transaction. AI fear marketing raises the threat, positions the company as the only entity capable of managing it, and uses that positioning to attract investment, regulatory goodwill, and consumer trust simultaneously. The fear does not just sell the product. It sells the narrative that the company is necessary.

We’re seeing the impact now. What makes AI unusual as a fear-marketing case study is that Witte’s two responses are happening simultaneously, across different segments of the same audience. One group is deep in danger control. They have accepted the threat as real, urgent, and personal — AI is coming for jobs, for relevance, for competitive advantage — and they are acting accordingly. Executives are mandating adoption. Developers are scrambling to upskill. Companies are embedding AI into workflows, driven less by enthusiasm than by the terror of being the ones who didn’t. This is the fear working as designed.

The other group has tipped into fear control. Overwhelmed by a threat that feels too large, too abstract, and too far outside their ability to meaningfully address, they have detached because the fear has no workable exit. You cannot buy protection from AI displacement the way you can buy an alarm system or open a savings account. The recommended action — adopt, adapt, upskill — is vague enough to stop you in your tracks. So the message gets dismissed, the doom discourse gets tuned out, and the scroll continues.

The AI industry has engineered the perfect fear-marketing machine: one product that simultaneously drives urgency in believers and numbness in skeptics, with both responses generating exactly what the companies need — adoption from one camp, and from the other, a public too exhausted to push back.