The Altman Bomb and the Reckoning They Refused

So here is Sam Altman, the man who spent May telling the world not to worry about a jobs apocalypse, spending June as the target of a suspected bombing plot. The Guardian reports that a 31-year-old was arrested outside OpenAI’s San Francisco offices with explosive devices and a manifesto. The suspect had allegedly threatened Altman’s life. The 'breakneck AI boom', as the paper puts it, is 'fueling anti-tech extremism'. No kidding.

You have to admire the symmetry. Now someone has decided that the appropriate response is not to write a letter to the editor, but to build a bomb. This is what happens when an entire industry gaslights the public for long enough: the polite scepticism curdles into something a great deal less polite.

And let us not pretend that this is an isolated incident by a lone crank. The researchers cited in the Guardian piece warn that AI is becoming 'a driver of political violence'. The YouGov data, reported in the FT, shows that more than 40% of Americans do not trust AI. That distrust is not abstract — it is tangible, and it is already shaping politics. When the technology’s cheerleaders treat every criticism as a failure of communication rather than a legitimate grievance, they should not be surprised when the grievance finds less communicative outlets.

The industry posture, of course, will be to double down on the security theatre. Altman will get more bodyguards. OpenAI will install more turnstiles. The conference circuit will produce breathless panels on 'addressing public concerns through transparency', which is Silicon Valley for 'we have noticed that people are angry and intend to use our words to make them less angry without actually changing anything'. But the problem is not a PR problem. It is a structural one. The same companies that insist AI will not cause mass unemployment are also laying off thousands of workers and citing AI as the reason. The same executives who speak of 'alignment' and 'safety' in their blog posts are racing to market with tools that automate jobs and centralise power. The gap between the rhetoric and the reality is no longer a crack — it is a canyon. And in that canyon, extremism grows.

There is a particular irony in the suspect targeting Altman himself. Altman is not the engineer who built the most destructive model, nor the venture capitalist who funded it, nor the politician who failed to regulate it. He is, however, its face — the affable, hoodie-wearing CEO who sells the future with a shrug. When your brand is 'I am the reasonable one, trust me', you become the target for everyone who has stopped trusting. The industry has created a class of celebrity executives who are at once omnipresent and evasive, promising utopia while delivering precarity. That is a dangerous combination.

Two weeks ago, in this column, I wrote that Altman's no-apocalypse gambit was a 'performance' — a bid to manage expectations while his customers continued to blame AI for their layoffs. I did not predict it would end with someone trying to blow him up. But let us be honest: the wonder is that it has taken this long. When you tell people their livelihoods are collateral damage in a revolution that will benefit 'everyone', and then assure them that no, actually, there will be no revolution, you erode the credibility of the entire system. And when credibility collapses, the people who feel left behind do not wait for the next earnings call. They act.

The industry will frame this as a security issue. It is not. It is a legitimacy crisis. The 'hurricane warning' that Politico ran the same day is not about model guardrails — it is about the human consequences of treating AI as an unstoppable force that needs only to be managed, not questioned. The guardrails that are 'too late' are not the ones inside the models. They are the social guardrails — the trust, the consent, the belief that change should be negotiated, not imposed. Those have been knocked down already. The bomb plot is just the signal flare.