(SeaPRwire) –
By: Silas Sterling
The online pile-on against Anthropic’s latest ad was immediate and loud. Sam Altman called it satire bait. A top YouTube comment labeled it dystopian marketing slop. Atlantic writer Lila Shroff called it tone-deaf. Most critics missed the ad’s actual goal. It wasn’t meant to sell Claude. It was meant to show Anthropic listens to public AI fears. The audio clips came from a U.S. roadshow last spring. The campaign drew from 120,000 surveys across 159 countries.
This isn’t the first time an Anthropic ad divided viewers. The company’s 2026 Super Bowl spot won a Cannes Lions Grand Prix. But it ranked in the bottom 3% of general audience likeability. Many viewers didn’t get the joke about OpenAI’s planned ChatGPT ads. That spot laid bare a wider cultural gap. Silicon Valley insiders live in a different bubble than the rest of the world.
The latest ad controversy points to a deeper rupture in public trust. When Anthropic pulled Mythos Preview over cybersecurity risks, governments took the warning seriously. Most members of the public did not. Critics called the move a sophisticated marketing stunt. That ignores a key detail. A hacker used a weaker Claude model to steal 150 gigabytes of data from Mexico two months prior. Independent tests show Mythos Preview is far more powerful.
Part of the public’s cynicism makes perfect sense. Anthropic is racing to build the same AI it warns about. The company’s logic is simple. If they stop building, others will keep going. So leading development is the safest path. That logic feels self-serving, but it’s not insincere. Anthropic has also joined OpenAI and DeepMind in calling for international coordination to slow model development. This mirrors the fossil fuel and tobacco industries. Those groups suppressed evidence of harm for decades. Being honest about risks is better than hiding them, even if it’s branded as PR.
The ad’s most provocative image was Arlington National Cemetery’s soldier tombs. Zack Korman, CEO of AI cybersecurity firm Embroidery, called the choice deeply problematic on X. Some critics called the pick distasteful and over the top. But for CEO Dario Amodei, the image was a stark reminder. He estimates AI carries a 10% to 25% chance of civilization-ending catastrophe. This ad was more honest than past Claude campaigns. Past ads framed Claude as a tool for tinkerers and problem solvers. The real failure here isn’t the ad itself. It’s that the public no longer trusts AI companies to act in their best interest, no matter their messaging.
Author bio: Silas Sterling, veteran kernel contributor and editor-in-chief of an open-source security digest.