It is not necessarily a bigger threat.  Using 'AI' to get some vulnerabilities out of a source code still means you have to know what to do with it.  A traditional hacker could simply discover a way more disastrous vulnerability than a machine could, so it depends how you measure the threat level.

We adapt or we do not.  Ideally, these 'super computers' will be gradually integrated in the hands of companies and then of average people.  If they can destroy a lot of things, they can destroy the Internet.  Many data bases, many servers, they could all have almost zero security in front of a 'super computer'.  And considering how much personal information sits in so many storage devices connected to the outside World, I would assume we will first have to have things ready before such a thing would reach our hands.  Other wise, that would be like handing out weapons to people before authorities even have shields against them.

AI can do good and harm and if the intention is to threat the security of any platform that includes bitcoin's network, it's possible. We've seen how many hacks have happened recently, in the crypto casinos and if it's combined with these clever minds of these hackers. It can be very threatening just like the other day, it was AI that have found the bug in php. So, we can't remove the fact that with all of these incidents, the use of ai and how quickly it evolves can be a threat not only to bitcoin but also for other systems that might have some critical flaw.

Either it's a threat or not, I know one thing. It is forcing every lazy developer out there to work on the projects they created but never worked on afterwards to make them hack-proof or at least more solid. Even BTC had two vulnerabilities, and now, with AI, what if someone finds another one? But that's a hypothetical statement.

Yes, AI is a big threat, and this post is not the first one to make anyone realize that. There have been a lot of hacks that have occurred because of AI usage, and I believe the models released, like Ollama, have contributed the most because online LLMs have a lot of restrictions that hackers could not get around.

But where AI is being used to hack, it is also being used by developers to make it better using the same technology. So after a few years, this phase will probably end, and the projects will come out of this phase better and more secure than before.

AI is merely just a tool in the right or wrong hand just we fear quantum computers would be, that's how AI is.
So any fears should be countered with better upgrades in knowledge and software system and more vigilance is needed because AI right now can be so real, you think it's a human.

Hackers didn't build AI, they only know how to manipulate it better to their taste and if developers can do same to counter this, then AI would be and serve us just fine and be of greater benefit than we think.

AI on its own is not threat to anybody, but its still the same traditional hackers who are using this AI tools to search for possible vulnerabilities around an online system. Because before the invention of sophisticated AI tools, hackers use to search for possible vulnerabilities around a system manually, of which since these AI tools got invented, it has made work much easier for hackers, which is why I guess the group of Bitcoin developers had to increase the security of Bitcoin core.