EU AI Act - Page 4
Stories from EU AI Act come first, newest first, then stories from connected topics (European Commission, GDPR, AI Regulation, Hub: AI Regulation, AI Policy & Safety, compliance), each labelled with where it comes from. This view is not indexed by search engines.
Latest
OFAC Sanctions Eight Canadian Money Services Businesses
Canadian MSB sanctions name eight entities under US Iran-related authority. OFAC lists six TRON addresses; FINTRAC confirms coordination.
FinCEN Orders US Transfer Restrictions on MEXC Global
MEXC FinCEN order bars covered US institutions from transfers to or from the exchange, starting 30 days after Federal Register publication.
FMF Jailbreak Rating Framework Opens for Comment to 21 Dec
FMF Jailbreak Rating Framework v1.0 rates cyber jailbreaks on four dimensions. Voluntary; comment open to 21 Dec 2026. Not law or certification.
EU Preliminarily Finds TikTok For You Feed Breaches DSA
TikTok DSA: Commission preliminarily finds For You rabbit holes breach DSA; design changes sought; fine up to 6% turnover only if confirmed.
Meta, Google and TikTok Sue Over California AB 1709
California AB 1709: Meta, Google/YouTube and TikTok sue over under-16 feed and autoplay bans; Act effective 1 Jan 2027; up to $50,000 per minor.
DOJ Launches Super Intelligence Task Force Under Woodward
Super Intelligence Task Force: DOJ forms DOJ-SITF under Woodward and Weiss, merging Bondi’s AI Task Force; public launch 5 Nov after an October memo.
METR Says Anthropic AI R&D Pace Is About Twice Historical
METR assessment: Anthropic AI R&D ~2x historical pace (uncertain 1.3–3x); AECI ~1.9x; ~80% AI-attributed is METR’s guess; heavy redactions.
OCC Gives Bank Examiners a Super Intelligence Risk Framework
The OCC super intelligence framework guides bank examiners on AI risk. It sets no new rules, and banks need no separate SI risk appetite, the OCC says.
OpenAI Bug Bounty Payouts Jump as Cathedral Joins Its Security Sprint
OpenAI has more than quadrupled typical bug bounty rewards, with up to $172,000 for exceptional finds, and brought in Cathedral for a security sprint.
Anthropic’s FlashProve Uses Cryptography to Prove Which Model Replied
Anthropic's open-source FlashProve prototype proves a reply came from a committed AI model without revealing its weights, at about 70-95x the GPU cost.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.
