Anthropic's Super Bowl ad campaign: brilliant strategy or playing dirty?

Summary

Anthropic's Super Bowl ad campaign is strategically brilliant but deliberately deceptive. The ads, created by agency Mother, depict AI assistants responding to user requests with increasingly obsequious behavior—pauses, flattery—before pivoting to product placements. One ad shows an AI fitness bot recommending height-enhancing shoe inserts after being asked for workout advice. The ads are shot vertically for social platforms and performing well, with one version called "Violation" accumulating nearly 6,000 likes shortly after launch.

How the deception works

Anthropics ads function as political attack ads that avoid naming ChatGPT directly while tarring the entire category of ad-integrated LLMs. This differs from comparable competitive campaigns grounded in truth. Apple's 2010 attack on Android's adult app store was factual. Bud Light's corn syrup campaign was verifiable. Mac's virus warnings reflected real consumer experience. Anthropic's ads intentionally create a false perception of how ad-supported language models will behave, exaggerating the sleaziness of the ad experience to scare consumers away from OpenAI's product when it launches. Claude itself defines "playing dirty" as achieving goals through tactics that are deceptive, unethical, or that violate understood rules and norms of a given context. The ads fit that definition.

Strategic payoffs

The campaign entertains, builds Anthropic's profile ahead of its IPO, strengthens relationships with researchers worth tens of millions to retain elite talent, continues the company's fear-based messaging around AI safety, and baits OpenAI into public response. Sam Altman switched to uppercase in his reply. The ads also increase public scrutiny of ad rollouts broadly, potentially creating regulatory headwinds in Washington.

The consumer trust risk

The campaign could backfire by damaging consumer trust in LLMs generally. People may become wary that they're being monetized without consent or that they can't trust AI outputs as genuine advice. This is particularly problematic because Anthropic has publicly stated it doesn't care about consumer adoption, yet the ad spend and strategic positioning suggest otherwise.

Anthropics claim that ad-supported models will make LLMs free for people who can't afford subscriptions rings hollow. The company has already lost the race to serve billions of users. Gemini and ChatGPT dominate consumer adoption. Anthropic cannot catch up at scale without ads, but positioning itself as the "clean" alternative makes it harder to launch ad products later without appearing hypocritical.

A stronger path

Anthropic could follow Apple's playbook instead. Apple carved out a defensible consumer position at smaller scale, roughly 1.5 billion active devices globally, by emphasizing emotional and principled differences like privacy, safety, and truthfulness rather than fear-mongering about competitors. Apple succeeded not by being first with smartphones but by building an ecosystem people wanted to be seen using. Anthropic could do the same with Claude by committing to genuine differentiation rather than deceptive messaging.

The irony cuts deep. A company built on constitutional AI principles and safety-conscious positioning is using advertising tactics that violate its own stated values. The ads work at driving attention and engagement, but only by misleading the public about a product that does not yet exist.