Not at all surprised. Even back in 2017 all these alt projects were hiring cheap devs and doing copycat scripts.
And also sleeper scripts were sold for years just hiding their loophole for years until enough money was in the system.
AI is now accelerating that. Truly bad situation.
AI had online models that had restrictions in place for such tasks, but with Ollama and offline models, this has become easier than before. But I agree with you that all these years, the same shit was being sold, and that has caused many of these exploits in the big DeFi names.
But every time, the reason was a little bit different. Like, the Ronin Bridge had no loophole. It was a social engineering attack. The Lazarus Group tried to run a bogus job interview, got their malware into the system, and hacked $600+ million in 2022. Many projects had some good people, but they left loopholes on purpose. Like in the Harmony Horizon project, a security researcher warned about a loophole around three months earlier, but they ignored it, their keys got compromised, and they lost around $100M in 2022. There are several events like this, but since the start of 2026, most bridge-related hacks have happened due to flaws in smart contracts because AI has helped hackers.
The DeFi sector alone lost $1.3 billion in 2026 haha. That's a lot of money, bro. Anyway, AI was being used by hackers to find exploits, but it is also being used by white-hat hackers, like that red team that secured many flaws in LN-related projects a few months back.