This website uses cookies

Read our Privacy policy and Terms of use for more information.

EU bans AI sexual-deepfake generators from 2 December — and two clocks stop on 13 November

CyberEyeQ Weekly — Issue #44 · 5 November 2026 · Covering 23 October – 5 November. The EU adds AI Act prohibitions that start on 2 December, Mexico and India deadlines land on 13 November, and DOJ books a record-size cyber False Claims settlement.

At a Glance

  • EU bans sexual-deepfake generators — AI Act prohibitions apply from 2 December 2026.

  • Mexico decree report due ~13 Nov — platforms must document child-safety compliance.

  • India Consent Manager rule 13 Nov — DPDP Rule 4 commences.

  • $13M cyber False Claims deal — ILC Dover; NIST SP 800-171 failures alleged.

  • FTC probes surveillance pricing — Section 6(b) orders to nine companies.

  • CISA 2015 protections lapse 11 Dec — industry presses for NDAA renewal.

Critical Actions

  1. EU AI Act: new prohibitions on sexual-deepfake and CSAM generators from 2 December (27 days). The AI Omnibus adds prohibitions on AI systems that generate non-consensual sexually explicit content and child sexual abuse material, and postpones high-risk rules to 2 December 2027. We relied on the Commission press release as summarised in our tracker; article-level scope is [Unverified]. Action: test whether any generative image or video feature can produce such content and document safeguards. European Commission IP/26/1714

  2. Mexico: child-safety decree report due about 13 November (8 days). The decree, published in the DOF on 14 October and in force on publication, requires reliable age verification beyond self-declaration, strictest default settings for new accounts, and a report of compliance actions within 30 calendar days. Reporting channel not confirmed. Action: assemble the compliance record and identify the receiving authority. Diario Oficial de la Federación

  3. India: Consent Manager registration mandatory 13 November (8 days). DPDP Rule 4 and Section 6(9) commence; Consent Managers must be registered with the Data Protection Board (Indian company, net worth at least ₹2 crore, interoperable platform, fit-and-proper directors). Penalty provisions follow 13 May 2027. Action: confirm any Consent Manager you rely on will be registered. MeitY

Enforcement Watch

  • DOJ: ILC Dover and ILC Astrospace, $13M — announced 3 November to resolve False Claims Act allegations that they billed the Army, Missile Defense Agency and NASA while not meeting contract cybersecurity requirements (no current system security plan, incomplete NIST SP 800-171 controls, no POA&Ms, CUI in a cloud lacking FedRAMP Moderate-equivalent security). Relator, the former cyber director, receives $2.35M. Allegations only. Action: verify your SPRS score and CUI cloud controls. DOJ

  • FTC: nine companies ordered to disclose surveillance-pricing data uses — 28 October Section 6(b) orders to firms including Walmart, Target, Delta, Uber and JetBlue on use of personal data to set individualized prices and worker pay. A study, not an enforcement finding. Action: inventory personalised pricing and pay models. FTC

  • FDA early alert: Medline beds and accessories — all Medline Brand beds removed or corrected over fire (electrical overheating) and entrapment risk; two deaths, 16 serious injuries as of 20 October; no class assigned yet. Action: identify affected beds and follow the firm's instructions. FDA

Deadline Watch

  • 5–6 Nov — CPPA Board meeting (GPC opt-out signal); Saudi SDAIA PDPL consultation closes.

  • ~13 Nov — Mexico decree report; India DPDP Rule 4; EDPB Guidelines 04/2026 comments close.

  • 15 Nov — China MOFCOM/MOST export-control catalogue (23 AI technologies) comments.

  • 16–17 Nov — California EO N-9-26 AI recommendations; FedRAMP RFC-0035 comments.

  • 20 Nov — EU Consumer Credit Directive 2 applies.

  • 27–28 Nov — FDA/CBER genome-editing guidance comments; EUDAMED legacy-device registration.

  • 30 Nov — Brazil ANPD sanctions-rules consultation; FDA CDRH guidance-agenda comments.

  • 2 Dec — EU AI Act generative-content prohibitions.

  • 5 Dec — FCA UK EMIR intragroup notifications (current process).

  • 7 Dec — FedRAMP CR26 VDR/VER; Treasury outbound-investment NPRM comments.

  • 10–11 Dec — Australia Children's Online Privacy Code; CISA 2015 protections lapse.

  • 28 Dec — ONC HTI-5 effective. 1 Jan 2027 — Greece age-15 minimum, US state laws, UK EMIR regime. 18 Jan 2027 — OCC GENIUS rule effective.

Around the World

  • Australia — High Court heard the under-16 ban challenge 3–4 Nov and reserved judgment; ban stays in force (penalties up to A$49.5M). High Court · eSafety

  • Türkiye — under-15 ban and e-Devlet age token live for new accounts since 1 Nov; existing accounts by 31 March 2027. Resmî Gazete

  • United States — OCC Bulletin 2026-54 (4 Nov) tells examiners not to treat advanced AI as a separate risk category (guidance, not a rule) OCC; ONC's HTI-5 final rule removes 14 certification criteria and the DSI model-card requirement, effective 28 Dec ONC; Treasury's outbound-investment NPRM would recast "AI systems" as "super intelligence systems", comments 7 Dec Federal Register; CPPA may advance mandatory-GPC rulemaking CPPA.

  • United Kingdom — Cyber Security and Resilience Bill cleared the Lords after two government defeats; back to the Commons Parliament. FCA PS26/21 finalises the UK EMIR intragroup regime from 1 Jan 2027 FCA.

  • European Union — EDPB draft Guidelines 04/2026 make fining the default response; comments to 13 Nov EDPB. Italy's NIS2 enforcement phase began 1 Nov ACN.

  • United States, cyber — a 35-organization coalition wants the 10-year CISA 2015 reauthorization kept in the FY2027 NDAA before protections lapse 11 Dec; CIRCIA's final rule remains at OIRA reginfo.gov.

  • China — export-control comments close 15 Nov; NDRC says it will accelerate foundational AI legislation (direction, not a draft) MOFCOM · CAC.

  • Cloud — first full FedRAMP 20x Class C certification granted 2 Nov FedRAMP.

❝

🔭 Signal Watch — EU Council backs the Digital Omnibus simplification of GDPR, ePrivacy and cyber reporting. On 21 October 2026 the Council of the EU agreed its general approach — its negotiating mandate — on the Commission's Digital Omnibus data package (procedure 2025/0360(COD)), a set of proposed simplifications to the GDPR, the ePrivacy cookie-consent regime and overlapping cyber-incident-reporting duties. A general approach is a procedural milestone, not law: the file moves into trilogue once the European Parliament fixes its own position, and nothing changes for compliance teams yet. It is, though, the clearest sign so far that the package will advance. What to do now: keep every current GDPR and ePrivacy obligation in force — none of them change until adopted text appears in the Official Journal — and start watching for the Parliament mandate and the opening of trilogues, since the Council text reaches cookie-consent rules and the consolidation of breach/incident-reporting notifications that most programs are built around. Source: Council of the EU press release, 21 Oct 2026.

Deep Dive: Cyber False Claims — what the $13M ILC Dover settlement says about NIST SP 800-171

The setup. The ILC Dover settlement is the latest in DOJ's Civil Cyber-Fraud Initiative, which treats a claim for payment as an implied certification that the contract's cybersecurity terms are met. It follows the $2,042,518 Honeywell Aerospace settlement of 1 September. The alleged failures are ordinary ones: no current system security plan, controls not fully implemented, no plans of action and milestones, and controlled unclassified information stored in a cloud outside the required baseline. What turned them into a False Claims case was a whistleblower, the company's own former cybersecurity director.

The analysis. The money matters less than the pattern. The government did not need a breach; the theory is that the paperwork told DoD one thing and the environment did another. For contractors that makes the self-assessment score the exposure point: a submission that is stale or optimistic is itself the alleged conduct. And with a $2.35M relator share, the incentive for insiders to raise the issue is plain. Here is where the audit of your own submission should start...

The analysis continues for CyberEyeQ Pro subscribers. Unlock the full deep-dive recommendations, all five weekly action items, source documents, and jurisdiction-specific compliance checklists.

logo

Subscribe to Pro to read the rest.

Become a paying subscriber of Pro to get access to this post and other subscriber-only content.

Upgrade

Already a subscriber? Sign in

A subscription gets you:

  • Deep dives of each updated regulation
  • Source verification documents
  • Extended jurisdiction-specific analysis
  • Compliance deadline tracker
  • Regulation crosswalk