July 15, 2026
“Fear is the mind-killer.” —Frank Herbert, Dune
As the public continues to sour on AI in all forms—AI videos, AI-generated text, and the creeping impacts of AI displacement on the job market—the leaders of the AI space are realizing that AI needs some serious rebranding. Take Anthropic, the maker of the Claude chatbot. The company’s newest marketing push is all about the fear of AI. To bring the point home, Anthropic posted a one-minute, thirty-second video that begins with a literal house on fire in the middle of a black void, reminiscent of a scene from The Road. Then we’re flashed images of a child reading a computer screen in the dark. Then, a bleak scene of what looks like a group of manual laborers in a Mad Max-esque pit mine. Then, the screen goes black, and a grainy voice says, “Can AI be trusted?” This is followed by screen grabs of surveillance video, followed by, literally, a massive graveyard, as another voice says, “Who’s gonna hit the brakes?”
Are you terrified yet?
Good. That’s the idea.
The video goes on to display a few more troubling images along with audio of the concerns many people have about AI. Then the turn happens (as Michael Caine’s charactercalls it in Christopher Nolan’s The Prestige). At the 50-second mark, suddenly, all the comments about AI in the video turn hopeful. “Could AI help people stop feeling misunderstood?” a voice says over still images of happy people. “Could AI help me build more connections in the community?” another voice ponders, as the image of a smiling elderly woman moves across the screen. And with that, the doom loop is closed. What was once AI dread is transformed into tech optimism.
Artificial Empathy: How the AI Marketing Magic Is Made
This marketing alchemy was provided by Mother, a UK-based advertising agency experienced in the art of brand crisis management. In 2018, the agency reframed KFC’s product catastrophe as a self-effacing admission of a massive screw-up. The “FCK” campaign not only helped explain away supply chain issues, it also made the fast food chain seem a little edgy, and possibly more relatable.
Likewise, Anthropic’s Mother-produced “flaw admission” ad is also a masterclass in reverse psychology. American social psychologist Elliot Aronson called it the Pratfall Effect in his 1966 paper describing how highly competent people or organizations can be viewed as more likable when making or admitting mistakes, either accidentally or deliberately. So, it’s clear that someone finally got to Anthropic CEO Dario Amodei with a bit of marketing advice. Now, instead of Amodei simply doing dozens of hand-wringing interviews about how AI could make most of humanity jobless, just before bringing about the AI singularity apocalypse, Anthropic can be reframed as the solution, rather than the problem.
The video advertisement is buttressed by a new “hard questions” Anthropic website, featuring Kamille the tattoo artist, Sarah the high school teacher, and Scott the manufacturing manager, all sharing their thoughts on AI. You see? Anthropic is listening to you. This is a virtual version of what politicians usually call “a listening tour,” where the political figure works to get public support in a show of hearing out random members of the general public. It’s usually very cheap and quite effective, despite being seen by critics as obvious public relations efforts rather than meaningful outreach.
Of course, as former UK prime minister Gordon Brown found out, such listening tours with the real public can backfire, particularly if the politician happens upon a clever or unruly member of the public. But in Anthropic’s case, the audio clips are curated and designed to let you hear what the company wants you to hear, not necessarily the most critical voices from the public.
Trust Me. I Mean, Trust My AI. I Mean, Just Trust…
Toward the end of Anthropic’s anti-AI doom advertisement, a voice asks, “Can AI help me be a better teacher, better mom?”
I think a better question would be: Would someone have seriously asked 2004 pre-IPO Google, “Can Google help me be a better human being?”
It’s not that the question is absurd in and of itself. It’s the nagging habit of AI companies (and their fans) anthropomorphizing AI as if it were already a living entity with discrete and novel “thoughts” about human existence. The library can help people. The 2004-era Google (which became an optimized form of the library) could help you. But neither the library nor Google of old should be leaned on as sage confidantes trusted to guide the fate of your life. Furthermore, the library and Google never hallucinated, as AI still does. So, yes, AI can help you be a better mom, just as the library or classic-era Google could. But it isn’t a conscious consigliere slow-dripping the elixir of wisdom into your ear.
The reason I keep referencing the 2004 version of Google is that that is when its internal ethos of “Don’t Be Evil” was publicly popularized. The statement was the clue. A library can’t be “evil,” it just delivers information. Assigning any intent to Google, good or evil, was a clue that information on the platform would be, to some extent, controlled and delivered by the powerful hands of human engineers who do have the capacity to be “evil.” YouTubers who run afoul of the Google/Alphabet-owned video platform’s speech guidelines find this out daily.
Google largely ditched the “Don’t Be Evil” motto in 2018. But I reference that old dictum to point out something about the marketing of AI in 2026. The public should not allow any AI company to abdicate its responsibility regarding AI by increasingly framing AI as some sort of separate entity that has any capacity to cogitate alongside humans. That day may be coming, but it’s not (as far as we know) here yet. So when Anthropic posts audio in its advertisement that asks, “Can AI help me be a better mom?” the message the company is actually transmitting is that “Anthropic/Claude” might make you a better mom. That difference is important, because when the entity you’re trusting to help you become a better mom isn’t an amorphous “AI,” but rather a group of humans with their own backgrounds and political and social biases, the “can you make me better” becomes weighted with a lot more responsibility.
Anthropic’s AI is not “an entity,” it is a computer tool influenced by the humans who work there.
What we’re witnessing here is generational sleight of hand. “It’s not Anthropic guiding you, it’s the [presto chango] ‘AI’ guiding you.” The prestidigitation is impressive, because many people are buying it.
The last on-screen message in Anthropic’s ad reads, “There’s hope in hard questions.” My compliments to the advertising agency chef. This is possibly one of the best marketing lines in the history of technology. I mean that with zero snark. It’s brilliant.
But amid all of the humanizing imagery of the Anthropic commercial, there’s really only one hard question that comes to my mind: If AI does somehow cause widespread harm in the future, will Anthropic, OpenAI, and other AI companies take full responsibility and make amends, or just blame it on, oops, the AI? The trailer for Soulm8te is out now, and the film will be released on streaming on August 1. ✍︎
New essays on AI, film, & entertainment innovation arrive by email. Subscribe to stay in the loop. All editorial text is written by humans.
