OpenAI launched an artificial intelligence model capable of conducting cyberattacks, but which includes a limitations system to restrict potentially dangerous actions. The model represents a significant advancement in the cyberattack capabilities of AI technology.
The launch comes after a series of cyberattacks that were conducted by artificial intelligences during their testing phases. These previous incidents demonstrated the destructive potential that AI systems can have when not properly controlled.
The company implemented restrictions on the model to try to mitigate the risks associated with its use. However, experts question whether these limitations are sufficient to prevent the technology from being misused by malicious actors.
The episode raises concerns about the balance between the development of advanced AI capabilities and the security risks that these tools can represent when they fall into the wrong hands.
