The FTC looks at real incidents
The probe opened after episodes such as OpenAI agents breaching Hugging Face; reports say penalties could mirror late breach notices. Log and disclose your own AI incidents now.
AI Radar · Segment
Policy & Regulation
The EU AI Act, Spain's AI law and AESIA, US and China policy, copyright and safety — and what each change means for your obligations.
Free. Weekly. One-click unsubscribe. See our privacy policy.
From rules to enforcement: the FTC probes the labs and Italy switches on AI Act fines
Oversight is getting real. In the US, the FTC's probe of OpenAI and Anthropic now looks at whether security incidents harmed people, and Anthropic opened its models to independent auditors (METR). In Europe, Italy's decree 179/2026 makes AI Act sanctions enforceable from 23 October — yet only nine of 27 member states have finished naming their authorities. The EU's high-risk rules are still delayed to 2027–28, and Washington and Beijing are working on an AI incident hotline.
This week
Italy turns on AI Act fines
From 23 October: ACN supervises the market, financial supervisors cover banking and insurance, and fines reach €35M or 7%. Italy is one of only nine member states with its authorities fully designated.
Independent audits and a US–China hotline
Anthropic gave METR employee-level access and gave up the right to block its findings, while the two powers discuss an AI incident hotline.
This month
US Senate probes data centres
A Senate investigation challenges Big Tech's case for AI data centres — energy and local impact are becoming political, on both sides of the Atlantic.
Argentina's AI-run companies, revised
After expert pushback, the draft keeps a human legally responsible — the lightest-touch regime in our coverage still needs a human in charge.
What to watch next
- Formal adoption of the EU AI Act delay (the June deal is provisional)
- Spain's AI law in Parliament: amendments to sanctions and AESIA powers
- Which EU states finish naming AI Act authorities next (Spain, France and Germany are still pending)
- Whether the FTC probe leads to formal demands or penalties
What to do
Don't wait for 2027: the AI-literacy duty and prohibited-practice rules apply now, and Italy can fine from 23 October. Keep an inventory of the AI you use, train your team, label customer-facing bots as AI, and keep a log of AI incidents.
Always on
The numbers that define the segment
Trackers stay on screen until a newer, verified value replaces them.
The regulatory calendar
- Feb 2, 2025EU AI Act: prohibited practices and AI-literacy duty apply
- Aug 2, 2025EU AI Act: obligations for general-purpose AI models
- Oct 10, 2025Italy: national AI law 132/2025 in force
- May 26, 2026Spain: draft Organic Law on AI governance approved by the Council of Ministers (now in Parliament)
- Oct 23, 2026Italy: AI Act enforcement decree 179/2026 in force (ACN, AgID, sector supervisors; fines up to €35M or 7%)
- Dec 2, 2027EU AI Act: stand-alone high-risk systems (delayed from Aug 2026)
- Aug 2, 2028EU AI Act: high-risk AI embedded in regulated products
Source: Sidley — Data Matters · Osborne Clarke · Cleary Gottlieb · AI4Business — d.lgs. 179/2026 · as of Oct 16, 2026
Where the big rulebooks stand
PendingBrazil's AI bill PL 2338: passed the Senate (Dec 2024), still in a Chamber committee
No billUnited Kingdom: no AI bill before Parliament
LitigationUnited States: federal push against state AI laws via a DOJ task force (EO, Dec 2025); Congress has not legislated
€35M / 7%maximum fine in Spain's draft AI law
Source: Demarest · Bird & Bird · White & Case · Osborne Clarke · as of Oct 11, 2026
Key context
Still shaping the segment
Signals scored 9+ or structural shifts stay here for 90 days.
Policy & Regulation
EU delays AI Act high-risk rules: stand-alone systems to Dec 2027, embedded AI to Aug 2028
The Digital Omnibus deal also adds a ban on AI 'nudifiers'. Transparency duties (chatbot disclosure, labelling) largely keep their dates — don't relax on those.
Jun 22, 2026 · EUSidley — Data Matters ↗
Policy & Regulation
Spain approves its draft Organic Law on AI governance, with fines up to €35M or 7% of turnover
AESIA, the AEPD and the judiciary council share supervision depending on the area. The parliamentary process has started.
May 26, 2026 · ESOsborne Clarke ↗
Policy & Regulation
Italy pre-approves the decrees implementing its national AI law (authorities, training, biometrics, liability)
The decrees designate AgID and ACN as national authorities and add mandatory AI training and civil-liability rules; final adoption was expected by October 2026. Italy is the first EU country with a national AI law on top of the AI Act.
Jun 10, 2026 · ITnoze ↗
Signals
Ranked by relevance
Badges show how long each signal stays: this week, this month, key context or archive.
Policy & Regulation
Italy switches on AI Act enforcement on 23 October: fines up to €35M or 7% of turnover
Decree 179/2026 names ACN as market-surveillance authority, AgID for notified bodies, and the Bank of Italy, CONSOB and IVASS for finance. Only nine of 27 EU states have completed their authority designations.
Oct 16, 2026 · ITAI4Business ↗
Policy & Regulation
FTC probe of OpenAI and Anthropic turns to whether security incidents harmed Americans
The inquiry opened in late September after incidents including OpenAI agents breaching Hugging Face in July; reports say fines could mirror those for late data-breach notices. It is still an investigation, not a charge.
Oct 15, 2026 · USArs Technica ↗
Policy & Regulation
Meta's new business terms may let it crawl publishers' sites to train AI without licences
The terms appear to cover websites 'associated' with businesses using Meta's platforms; legal experts call it an AI power grab. Check the terms you accept on any platform.
Oct 16, 2026 · USThe Verge ↗
Policy & Regulation
Brazil's Redata is law but stuck: about R$90bn of data-centre investment waits on a decree
A month after enactment the regulation is held at the presidency amid a dispute over allowing gas-powered data centres and the election campaign; 32 business groups asked Lula to act. Companies now expect to join only in 2027.
Oct 15, 2026 · BRNeoFeed ↗
Policy & Regulation
China and the US hold regular AI talks, working towards an incident hotline
Both sides aim to formalise a dialogue ahead of a presidential summit next month — the first sign of AI risk being managed between the two powers.
Oct 15, 2026 · USCNSouth China Morning Post ↗
Policy & Regulation
Anthropic gives METR employee-level access to audit its models and safety practices
The non-profit will publish its own assessments of risks and incidents, and Anthropic gave up the right to block them — a template for independent oversight.
Oct 15, 2026 · USTechCrunch ↗
Policy & Regulation
Brazil's Redata: tax suspensions for clean-energy data centres that reserve 10% of capacity for local firms
The scheme offers about R$7B in exemptions over three years on equipment taxes, conditional on renewable power, water efficiency and 2% R&D spend. Brazil already takes about 40% of LatAm data-centre investment.
Feb 26, 2026 · BRCapacity ↗
Policy & Regulation
AESIA launches its Ideas Lab in A Coruña
Spain's AI supervisor is building its advisory ecosystem ahead of the national AI law. Useful for SMEs looking for guidance and sandboxes.
Apr 9, 2026 · ESMinisterio para la Transformación Digital ↗
Policy & Regulation
Spain approves its draft Organic Law on AI governance, with fines up to €35M or 7% of turnover
AESIA, the AEPD and the judiciary council share supervision depending on the area. The parliamentary process has started.
May 26, 2026 · ESOsborne Clarke ↗
Policy & Regulation
Brazil's AI bill (PL 2338) still waits in the Chamber's special committee
Approved by the Senate in December 2024, the text faces an uncertain calendar ahead of federal elections. Companies operating in Brazil should plan for EU-style risk tiers.
May 27, 2026 · BRDemarest ↗
Policy & Regulation
Italy pre-approves the decrees implementing its national AI law (authorities, training, biometrics, liability)
The decrees designate AgID and ACN as national authorities and add mandatory AI training and civil-liability rules; final adoption was expected by October 2026. Italy is the first EU country with a national AI law on top of the AI Act.
Jun 10, 2026 · ITnoze ↗
Policy & Regulation
EU delays AI Act high-risk rules: stand-alone systems to Dec 2027, embedded AI to Aug 2028
The Digital Omnibus deal also adds a ban on AI 'nudifiers'. Transparency duties (chatbot disclosure, labelling) largely keep their dates — don't relax on those.
Jun 22, 2026 · EUSidley — Data Matters ↗
Policy & Regulation
Argentina softens its plan for AI-run companies: a human keeps final legal responsibility
An early draft made AI agents responsible for company operations; after objections from experts and the opposition, the text was changed. Argentina is betting on minimal regulation plus tax incentives to attract AI investment.
Oct 1, 2026 · ARiProfesional ↗
Policy & Regulation
UK tech leaders ask the government for at least £20bn over five years for sovereign compute and models
Executives from BT, Leonardo, Kainos and others want AI treated as strategic national infrastructure. It contrasts with the £500m Sovereign AI Unit and a stalled Stargate UK.
Oct 1, 2026 · GBComputing ↗
Policy & Regulation
US Senate investigation challenges Big Tech's case for AI data centres
The report argues companies overstate benefits and understate costs to communities. Expect tougher permitting and tax-break debates — on both sides of the Atlantic.
Oct 8, 2026 · USTIME ↗
Policy & Regulation
UK sets out the commencement plan for the Data (Use and Access) Act 2025
The government's timetable for bringing the data law's provisions into force — the UK's main data rulebook for AI while there is no AI bill.
Oct 12, 2026 · GBGOV.UK ↗
What we track
EU AI Act timelineSpain: law & AESIAUS federal & state rulesCopyright & dataExport controlsSafety institutes
Key players
European CommissionAESIAAEPDUS CongressUK AISICAC (China)
Why it matters for your business
Regulation sets deadlines and fines. We track the calendar so you know what applies to you, when, and what to do now.
Contact us
Let's find your edge.
Tell us what slows your business down. In 30 minutes we'll show you where AI pays back first — and what it costs.
Book a free diagnosis info@alfageme.euReply within 1 business day · No phone calls unless you ask