The EU's AI Act amendments landed in the Official Journal on Friday and enter into force today. High-risk obligations slip from 2 August 2026 to December 2027, and to August 2028 for AI built into regulated products. The same regulation bans AI that generates child sexual abuse material or non-consensual intimate imagery, from 2 December 2026. Those systems carry the same risk they did last week, and teams building for 2 August now have to justify that work without a regulator's clock behind it.