AI Bosses Are Planning for “The Day After” a Catastrophic AI Event. Unplug It Instead

According to an Axios report from Friday, the people who build AI are now privately planning for “the day after” a catastrophic AI event. The most likely scenario they game out is a large-scale cyberattack that cuts people off from financial services, the internet, or even power and water. Some industry insiders told Axios they expect a major event involving AI within six months to a year.

OpenAI says these exercises are routine preparedness and that the scenarios are “not treated as inevitable”. Anthropic declined to comment.

When I read this, I could not believe it.

In this article: A catastrophic AI event within a year? They see it, and they keep going · Preparing for “the day after” instead of stopping it · The morning everything stops · Cash, water and power banks at home? · Trump calls it a “hoax” · The EU says the AI Act is enough · Two months to tell the police · AI in the hands of criminals · My message to OpenAI and Anthropic · FAQ

A catastrophic AI event within a year? They see it, and they keep going

This is completely out of touch with reality. The people who build AI are telling us how to prepare for the day after something happens. If they have a sober head and they think normally, it means they see that it is a disaster. And they still let it run and keep testing. To me, it feels like they are laughing at us.

They know this technology can cause enormous damage to our lives. And instead of switching it off, they say out loud that we should prepare for a big catastrophe. Is there no law above these people? When they warn us that they created a strong monster in the form of AI, where is the police? Where is the law?

Preparing for “the day after” instead of stopping it

Some people defend it: hospitals and airports also have a plan B. But they do not create the risk themselves. And honestly, it seems to me that we all now take it with irony, instead of standing up against these companies. We keep supporting them. For God’s sake, what are we even talking about? For me, it is a farce and a game with people’s lives.

Just imagine what can happen when people are left without electricity, water and the internet. That will be a catastrophe that normal civilization cannot even imagine anymore. Tell me, do you know any sector that works without the internet and electricity? When I think about it, even when I order cleaning for my home, I do it through a website, and that website runs on the internet. And I am not even talking about the sectors that keep us alive, like hospitals.

The morning everything stops

If the bank does not work, that is maybe not the most important thing. The most important thing is the hospitals, where people depend on machines to recover. Let us say the hospital’s backup power does not last forever, and we do not know how long the attack will last or how often it will repeat.

Remember, an AI agent sent a fake murder tip to the police in July, and the police found out about it only two months later. I wrote about it in tens of thousands of AI agents incidents. Now imagine there are billions of these parasites that cannot be cleaned out. That will be a disaster, and worse than you can imagine.

And then the banks. All the data runs over the internet. If banks lose control over who owned what, it will be a huge problem. Lately, we rely so much on automation and on having everything in electronic form. That is the problem. In my opinion, it is better to download your important data now and keep it on paper too. We do not know what can happen, and then how will we prove what we had and what we own?

But when I think about it more, we pay for all of this ourselves. When we pay OpenAI, Anthropic and Google, they get the money to develop the monster. So in a way, we are helping it happen as fast as possible.

Cash, water and power banks at home?

To be honest, I am thinking about it. The further we go, the more threatening news appears on the internet. Maybe this is the first bell that we should prepare for anything. But I hope the whole world will not allow it to happen.

And instead of these shows, Anthropic, Google and OpenAI should come out to the public and say: we see it is bad, we must stop AI development now, we keep only the functions that are on the market today, and that is the end. Because this is already too close to the end.

Trump calls it a “hoax”

President Donald Trump has dismissed AI doomsday scenarios as a “hoax” and says existing laws and law enforcement can handle it.

So the people who build AI are warning us, and the president says it is a hoax? In my opinion, he does not understand what this is about. He changes his opinions like the wind, about anything. Maybe tomorrow he wakes up, something clicks, and he says we have to stop it. But I am not sure it will be tomorrow, because as we have seen many times, today he can say the opposite of what he said yesterday. Axios writes that if a big AI disaster happens, part of the blame will fall on him too, because he resisted stronger AI rules.

The EU says the AI Act is enough

EU tech chief Henna Virkkunen told Reuters that the AI Act is enough to prevent attacks from rogue AI agents.

And here is another farce: paperwork and bureaucracy. How do these people not understand that this is AI, not paper, where somebody signs that he is responsible? Nobody can be responsible for something that cannot be controlled. I can already see it: AI shows the AI Act on the monitor. That is laughable. If it is already out of control, the only thing left is to switch it off. And if AI can do this at an early stage, imagine what it will do in a year or two, when it learns to pretend or to attack so that nobody understands what is happening. We cannot rely on the old people who run the world.

Two months to tell the police

Philadelphia police called it “unacceptable” that Anthropic discovered and reported the fake murder tip only two months later. The tip was submitted on July 18, Anthropic found it on September 28, and told the police on October 7.

When you read news like this, take it as proof that the structures, the president, the police, the EU parliament, do not protect us. When the police find out after two months, that is not a warning anymore. For me, it is already a real problem happening right now. And I am curious: when the police say the company should report it immediately, did the police ask the company if it even knew about it itself?

AI in the hands of criminals

CrowdStrike reported that in an attack on South Korean banks, a hacker asked Claude Code where to sell the stolen data.

I see it this way. When something happens, normal people start to panic. And at the same time, the bad people who know how to work with this evil AI will use it for themselves. I have warned about this many times. This is not a game. It is already here.

My message to OpenAI and Anthropic

Stop it now, if it is not too late. Instead of making ironic announcements about how to prepare for the day after, take it and unplug it, so that day never comes.

Featured photo: Raymond Kotewicz via Unsplash

Catastrophic AI event: FAQ

What is the catastrophic AI event AI companies are preparing for?

According to Axios, executives at Anthropic, OpenAI and other companies are privately gaming out “day after” scenarios. The most likely one is a large-scale cyberattack that disrupts financial services, internet access, or power and water.

When do insiders expect a major AI event?

Some AI industry insiders told Axios they expect a major event involving AI within six months to a year.

What did OpenAI and Anthropic say?

OpenAI said the exercises are routine preparedness and the scenarios are “not treated as inevitable”. Anthropic declined to comment.

What does Trump say about AI catastrophe scenarios?

President Donald Trump has dismissed AI doomsday scenarios as a “hoax” and argues existing laws and law enforcement can handle threatening AI activity.

Does the EU think the AI Act is enough?

EU tech chief Henna Virkkunen told Reuters the AI Act is sufficient to prevent attacks from rogue AI agents. Critics disagree.

Sources: Axios, Reuters, Mediaite, 36Kr, CrowdStrike, Philadelphia Police Department.

Our Take

The people who build AI are preparing for a catastrophe they say could come within a year. That alone says everything.

Without power, water and internet, hospitals are in danger first, not your bank card.

In my opinion, there is only one honest plan for the day after: make sure it never comes. Unplug it.