Overview
- An internal draft reported by the Financial Times proposes a one-year grace period before enforcing rules on existing high-risk AI deployments.
- Fines for transparency violations would be postponed until August 2027 to allow companies time to adapt, according to the draft.
- Possible tweaks include looser post-market monitoring for high-risk systems through less prescriptive guidance, as reported by MLex.
- A package of simplification measures is expected around November 19, and any changes would require approval from EU member states and the European Parliament.
- Pressure on Brussels includes pushback from the Trump administration and an open letter from 46 European companies calling for a two-year pause.