Italy forces DeepSeek, Mistral and Nova AI to warn users about hallucinations.
The Story
Italy's regulator issued warnings requiring these models to disclose the potential for generating incorrect information.
Why It Matters
This is a real-world preview of the EU AI Act's transparency rules. For pharma, legal, or finance firms, using AI that hallucinates means real operational risk. Your team needs to understand that AI outputs are not facts. Treating AI as an oracle for critical tasks will create liability.
What To Do About It
Review your AI use cases this week. What critical decisions are based on AI-generated content? Implement a human-in-the-loop verification step for any AI output used in client-facing work or regulated processes. I've seen this before; trust, but verify.
Sources
Related Signals
11 sierpnia weszła w życie polska ustawa o systemach sztucznej inteligencji, tworząca krajowy organ nadzoru KRiBSI z uprawnieniami do kontroli, żądania dostępu do systemów i nakładania kar.
19 Aug 2026
Brussels began enforcing the EU AI Act's transparency rules on 2 August 2026, with fines of up to 15 million euro or 3 percent of turnover now on the table.
13 Aug 2026Od 2 sierpnia 2026 r.
13 Aug 2026