A few days ago, I had an interview at a company. I was told that they don't use AI, they don't use any kind of AI tool, they have critical operations of the Diorps, and any kind of negligence in any operation can lead to losing your job.
And as far as AI replacing hackers and penetration testers, it doesn't seem to be anywhere near as good as it could be.The picture is very blurry and it is becoming increasingly difficult to predict anything within such a blurry picture. The most important thing is that AI must first return the money that was taken from the investors.
Four days ago, a new kind of spectacle began on the Internet: the most critical CVE of 2026, (cve-2026–41940) which affected 70 million websites.
This is very surprising and we are going to see how far it goes. A few days ago I was searching for the same thing, what would happen if a critical vulnerability was discovered within the SSH.
Whether the world is destroyed by an atomic bomb or not, one thing has become true: the world will definitely be destroyed by these new CVEs.
AI is creating many problems in many places, from writing code to court security. AI is fixing one security issue and leaving three or four new loopholes behind.
I also realized something after four days. For four days, I was preparing for an interview. I was applying for a network administrator position. At one place, I was offered a Linux administrator position. At another place, another company offered me a job. Then I was worried. But today, the internet opened up all over the world on Twitter.
After seeing all these things, I also found an account on Twitter where Operation AI Shutdown was written. This thing was clearly telling that in the coming times, a lot is going to happen against AI. Whether AI works or not, but the attacks on AI have become more than all the possibilities because hackers are now targeting LLM models in many places.
A few days ago, all these things were proven when a couple of months ago, a famous cloud company's data was breached and according to a report, all that data was passed through AI. They used that AI to engage in internal attacks.
Do you know what kind of program can perform prompt injection? These peaceful injections are still available in many places, in which you can convert the system into a kernel shell and attack from within the sandbox.