When the CEOs selling AI started warning it would kill us all, the fear got harder to take seriously

Key points

• A new cartoon by First Dog on the Moon, published by The Guardian, satirises the AI existential-risk debate.
• The comic's central irony: tech company CEOs are now among the loudest voices warning that AI could destroy humanity.
• The piece contains no research findings or peer-reviewed evidence and should be read as opinion and satire.
• Prints are sold through the First Dog shop.

There is a real tension at the centre of the AI safety conversation, and a cartoonist has put his finger on it more cleanly than most analysts have.

First Dog on the Moon, the pen name of Australian cartoonist Andrew Marlton, published a new comic this week through The Guardian. The premise is simple: for a long time, the people most worried that artificial intelligence, software systems capable of learning and making decisions without explicit human instruction, might end civilisation were researchers at the fringes of mainstream science. Then the CEOs of the biggest AI companies started saying the same thing. Somehow that made it feel less alarming, not more.

The joke writes itself, but the underlying point deserves a moment. When the people building and selling a technology also claim it could kill everyone, two readings are available. One: the danger is genuine and they deserve credit for candour. Two: apocalyptic language serves a purpose, making the product sound consequential, crowding out smaller competitors who cannot claim the same scale of risk, and positioning the speaker as the responsible adult in the room. Neither reading is automatically correct. Both are worth holding at once.

We've tracked this pattern across fourteen existential-risk stories since 22 August 2026, including a September report finding that doomsday predictions have become a recurring industry feature without ever slowing the race. The cartoon doesn't settle those questions. It punctures the performance of concern when concern is being performed by people with an obvious financial interest in the drama.

Should you change what you believe about AI risk?

No single cartoon should move that needle. Credible scientists disagree, the evidence base is still developing, and the loudness of any warning tells you very little about its accuracy. Who funds the warning, and what they stand to gain, are always fair questions. Our earlier piece "AI Could Kill Us All, Some Experts Warn. Should You Believe Them?" laid out what researchers on both sides actually said, which is a more useful starting point than a CEO's press release.

The cartoon is a needle, not a conclusion. It's doing exactly what good satire should, and the fact that it needs doing at all says something about how the safety debate has drifted from the lab into the marketing department.