Donald Trump posted an AI-generated image to Truth Social on Sunday depicting himself bathed in divine light, wearing robes, healing a man in a hospital bed while a demon hovered in the background. It looked exactly like what it was: an AI rendition of the President of the United States as Jesus Christ.

Then his own base told him to take it down. And he did.

The timing was not subtle. Hours before posting the image, Trump had gone after Pope Leo XIV on Truth Social, calling him "weak" for opposing the Iran war. So the sequence reads: attack the Pope, then post yourself as Jesus. Even for Trump, this was a lot.

Religious conservatives who have spent years defending Trump against accusations of blasphemy could not defend this one. Christian commentators called it "gross blasphemy." Franklin Graham, who has been one of Trump's most reliable evangelical allies, publicly urged the president to remove it. He did, on Monday morning.

Trump's explanation? He told reporters he thought it was an image of him as a doctor, not Jesus. The New York Times reported this with a straight face. The image showed him in flowing robes with divine light emanating from his hands while a literal demon floated behind the patient. Nobody on earth would mistake that for a medical professional.

The AI Angle Nobody Is Talking About

Here is what the political press is missing while they debate the theology: this is the first time a sitting US president has been caught sharing AI-generated propaganda about himself. Not deepfaked by an opponent. Not fabricated by a foreign troll farm. Created or sourced by Trump's own team, shared by Trump's own account, depicting the president as a literal deity.

We have spent two years debating whether AI deepfakes will be used to attack politicians. We completely missed the possibility that politicians would use AI to deify themselves.

Think about what just happened. A head of state used a generative AI tool to create religious iconography of himself and then shared it with millions of followers during an active military conflict and a public feud with the head of the Catholic Church. This is not a hypothetical scenario from an AI ethics textbook. It happened this morning.

What This Means

The guardrails conversation is wrong. Every AI safety debate has focused on preventing bad actors from creating deepfakes of politicians. Nobody built guardrails for when the politician is the one creating the deepfake of himself as God.

Every major AI model has filters to prevent generating images of real political figures in compromising situations. But none of them have filters for "the subject requested the image." Trump's team either generated this image using an AI tool that allows political figures, or they found one without those guardrails. Either way, the safety framework has a blind spot the size of the Oval Office.

The backlash worked this time. Trump deleted it. But the precedent is set. If AI-generated self-deification is now in the political playbook, it will not stay deleted for long.

First reported by Politico. Additional reporting from Reuters, The Guardian, BBC, CNBC, and The New York Times.