Anthropic’s ‘Keep Thinking’ Campaign: AI Responsibility or Just Another Summer PR Wash?

Let’s not kid ourselves: it’s Saturday, the city’s boiling, and Anthropic wants you to stop scrolling TikTok and watch their latest ‘Keep Thinking’ film about AI responsibility. They’re not doing this out of sheer civic duty. This is pure, high-gloss PR—engineered to make you feel like Claude’s existential crisis is your bedtime story.

The film, if you actually bother to sit through it between rooftop drinks and dodging e-bike delivery guys, is the latest in Anthropic’s semi-annual chest-thumping about ‘hard questions’ in AI. They posture like Socrates with a cloud GPU, except the questions are so sanitized they could double as a Google SafeSearch result. The real hard question—who profits when LLMs hallucinate their way through your data—doesn’t get a cameo.

Anthropic’s campaign is a masterclass in Silicon Valley guilt-washing. They want to look thoughtful without actually conceding that AI’s biggest risks aren’t theoretical—they’re already here. Ask anyone who’s had their site scraped into an LLM training set without consent, or who’s watched generic AI content flood Google’s index while Mountain View shrugs. The ‘responsibility’ they’re talking about is mostly about optics, not outcomes.

Here’s the uncomfortable truth: if you’re shipping AI products in 2026 and your idea of responsibility is a moody film and a hashtag, you’re part of the problem. Real responsibility means shipping verifiable transparency, audit logs, and tools that let users see—and control—what’s being done with their data, in real time. Anything less is just a summer rerun.