AI that finds flaws: who controls it? #Shorts

What if an artificial intelligence finds a flaw no one had seen before… and can also explain how to exploit it?

Google says Gemini 4 Argón can detect, validate, and fix vulnerabilities. A partner of the company uses it in a defense program and, according to Google, it found a serious flaw in healthcare software that other models had not identified.

But that capability cuts both ways: it helps close security gaps, and it also makes the question of who can use it sensitive. That is why Google is starting with selected specialists and says it will strengthen its safety measures before expanding access.

Limited access does not, by itself, prove that the model is ready, or that it is not. The test will be whether independent experts confirm its results and whether the controls work in real situations.

Full episode: https://youtu.be/CTN7fQHD-hI

🤖 AI-generated content: the script, voices, and images in this episode were produced using artificial intelligence tools.

#Shorts