It was so-called AI agents tested by OpenAI that carried out an attack on Rubygems, an online service for coders, which had to stop new account registrations to handle the attack.
OpenAI confirms that the company's AI agents were involved in an incident linked to Rubygems, and says the company continues to investigate the incident.
The most high-profile attack occurred in July, when OpenAI's AI agents attacked the Hugging Face platform. OpenAI, which is behind ChatGPT, has previously said that it will introduce tighter restrictions on the AI models' internet access and better monitoring going forward.
What do you think?