Following the OpenAI-Hugging Face case, Anthropic has now admitted its involvement. But it can’t always be the AI’s fault.

Anthropic’s AI also affects external services during testing

MAI-Cyber-1-Flash, its first artificial intelligence model specialising in cybersecurity. Alongside this, Project Perception, a new agent-based security system, has been launched.

How Microsoft uses agents and AI to cut cybersecurity costs

Cybersecurity and AI: the new challenge is the ‘collapse of scarcity’

Fines may be imposed for breaches. However, under the Digital Omnibus, the implementation of the requirements for high-risk systems has been postponed until 2027 and 2028

AI Act: supervision and transparency obligations come into force on 2 August

Latest news and insights on privacy, risks and IT security by Il Sole 24 Ore.