Meta-observatory · operational monitoring

Policy, privacy and AI observatory

We bring order to observatories, authorities, standards hubs and repositories: we catalogue sources, normalize their role and read convergence, divergence and blind spots before translating them into operational signals.

Existing registry verified 17 August 2026 · AI Observatory catalogued with methodology under source review · no automatic ingestion.

Explore the meta layer Source registry August update Operational monitoring Agentic incidents

Next reviewEDPB opens consultation on anonymisation and web scraping for generative AI

D-17 · 30 Oct 2026 · EDPB

Sources catalogued8Normalized observatories, authorities and hubs

Evidence-ready7Verified to contribute to signals

Source review1Catalogued, but excluded from operational conclusions

Cross-source lenses3Convergence · divergence · blind spot

Evidence gateCatalogued ≠ verified ≠ evidence-ready

Meta layerCross-source inference, never single-source synthesis

Operational layer1 future reviews · 5 high priority

Meta-observatory

Observatory registry

Not another feed, but a reasoned map of the ecosystem: who provides context, who defines obligations, who enforces them and who helps implement them. The insights below are inferences from the catalogue, not independent regulatory facts.

01Catalogue
Identify relevant observatories, authorities, standards hubs and repositories.
02Normalize
Separate source class, region, evidence role and verification status.
03Triangulate
Read relationships and gaps across independent sources without flattening them.
04Translate
Move only human-reviewed, evidence-gated material into the watch board.

01Convergence

From principle to proof

The census connects policy context, binding implementation and enforcement with standards and implementation support. The catalog-level signal is the relationship between these layers; it is not a claim that every source reaches the same conclusion.

ImplicationRead a policy theme across authority, enforcement and implementation lenses before turning it into an operational claim.

Sources compared

OECD.AI EDPB EU AI Office FTC UK ICO NIST AIRC IEEE ISoPE

02Divergence

Different clocks, different authority

Regulatory obligations, consultations, standards work and observatory analysis move at different cadences and carry different authority. A single undifferentiated feed would hide those differences.

ImplicationKeep dates, authority and evidence role attached to every update; compare them without flattening them.

Sources compared

OECD.AI EDPB EU AI Office FTC UK ICO NIST AIRC IEEE ISoPE

03Blind spot

Coverage is not yet global

PolicyWatcher's present census includes Global, EU, US and UK entries, but no dedicated authority source for Africa, Latin America or Asia-Pacific. This is a gap in our catalog, not evidence that those regions lack governance activity. AI Observatory is catalogued under source review and cannot close or substantiate those gaps.

ImplicationPrioritize authority-source discovery in the uncovered regions and keep new observatories behind the evidence gate until reviewed.

Sources compared

OECD.AI EDPB EU AI Office FTC UK ICO NIST AIRC IEEE ISoPE AI Observatory · review

Trust rule: no single-source synthesis. Sources under source review remain in the catalogue, but do not contribute to operational conclusions.

Source census · evidence registry

Source types and evidence thresholds

The registry makes different sources comparable without declaring them equivalent. AI Observatory is included with Source review status: visible, but not evidence-ready.

Source

OECD.AI Policy Observatory

Country policy resources and AI policy analysis for cross-market context.

Intergovernmental policy observatory · 17 August 2026

AI governanceRegulatory updatesEvents

ClassObservatory

RegionGlobal

Evidence rolePolicy context

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

EDPB News

Board news, coordinated enforcement updates and guidance signals.

EU data protection body · 17 August 2026

Privacy enforcementRegulatory updatesEvents

ClassAuthority / regulator

RegionEuropean Union

Evidence roleEnforcement

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

European AI Office

AI Act implementation, model governance and institutional coordination.

European Commission AI governance office · 17 August 2026

AI governanceRegulatory updatesEvents

ClassAuthority / regulator

RegionEuropean Union

Evidence roleBinding implementation

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

FTC Technology Blog and News

Consumer protection, privacy, competition and technology enforcement context.

US consumer protection agency · 17 August 2026

Privacy enforcementAI governanceRegulatory updates

ClassAuthority / regulator

RegionUnited States

Evidence roleEnforcement

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

UK ICO News

Regulatory news, consultation updates and enforcement context for UK data rights.

UK information rights regulator · 17 August 2026

Privacy enforcementRegulatory updatesEvents

ClassAuthority / regulator

RegionUnited Kingdom

Evidence roleEnforcement

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

NIST AI Resource Center

AI RMF, standards resources and implementation support materials.

US standards and measurement institute · 17 August 2026

StandardsAI governanceEvents

ClassStandards / implementation hub

RegionUnited States

Evidence roleStandards and implementation

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

IEEE ISoPE

Standards-oriented events and policy engineering references.

Standards and policy engineering community · 17 August 2026

StandardsEventsAI governance

ClassStandards / implementation hub

RegionGlobal

Evidence roleStandards and implementation

Evidence gateVerifiedEvidence-ready

AccessPublic web

Source

AI Observatory

Its methodology and outputs must pass the PolicyWatcher source gate before they can generate operational signals. It is not used as sole evidence for substantive regulatory claims.

Independent/research observatory · Catalogued 19 August 2026 · methodology pending

AI governance

ClassObservatory

RegionGlobal

Evidence roleResearch and context

Evidence gateSource reviewNot evidence-ready

AccessPublic web

Applicability update · verified 17 August 2026

EU AI Act updates

These are applicable obligations and enforcement milestones, not proposals under monitoring. Every entry links to the European Commission source used for verification.

01In force 2 Aug 2026

In force

AI Act Article 50 transparency obligations are now applicable

From 2 August, providers must support machine-readable marking of generated or manipulated content and deployers must disclose defined AI interactions or synthetic content. The 2 December 2026 window applies only to marking and detection for systems already on the market before 2 August.

EU AI Office

02Enforcement from 2 Aug 2026

In force

Full Commission enforcement begins for GPAI obligations

On 2 August the Commission moved from the collaborative phase to full enforcement of general-purpose AI model provider obligations, including additional measures for models with systemic risk.

EU AI Office

03Supervision from 3 Aug 2026

In force

Supervision and enforcement of AI literacy measures begins

From 3 August authorities supervise and enforce the AI literacy measures already required by the AI Act. This entry reflects the framework in force and does not repeat superseded wording from the earlier Omnibus proposal.

EU AI Office

Active monitoring

Review calendar

Upcoming reviews are ordered by date; overdue items remain visible with source context and the required action.

EDPB opens consultation on anonymisation and web scraping for generative AI

The 8 July guidelines address purpose limitation, transparency, data minimisation and special-category data; consultation closes 30 October 2026. The EDPB also adopted final blockchain guidance.

Source: EDPB
Category: Privacy enforcement
Region: European Union
Review: Reviewed during EU privacy sweeps

30 Oct 2026Consultation closes · 09:00 UTC review

Source

02Review due

ScheduledEvent

European AI Office review window

Recurring review window for AI Act implementation context and model-governance materials.

Source: EU AI Office
Category: Events
Region: European Union
Review: Reviewed during EU AI Act monitoring

First week monthlyManual review

Source Add to calendar

03Review due

ScheduledEvent

IEEE ISoPE 2026 programme review

Review public IEEE ISoPE materials before the October 1-2, 2026 privacy symposium window in New York.

Source: IEEE ISoPE
Category: Events
Region: Global
Review: Reviewed during standards and events sweeps

16 Sep 202609:00 UTC

Source Add to calendar

04Review due

Active monitoringGuidance

ICO sets out its 2026/27 safe AI workplan

The 29 May response identifies forward work on the AI code, agentic-AI guidance and consumer support. It is a workplan signal, not completed guidance.

Source: UK ICO
Category: AI governance
Region: United Kingdom
Review: Reviewed during UK privacy sweeps

01 Sept 2026PolicyWatcher review · 09:00 UTC

Source

05Review due

Active monitoringGuidance

FTC enforcement page lists AI accuracy request for comment

The current privacy and security enforcement page lists a 1 July request for comment on a proposed AI accuracy policy statement. PolicyWatcher treats it as US monitoring context, not a final rule.

Source: FTC
Category: Regulatory updates
Region: United States
Review: Reviewed during US enforcement sweeps

14 Aug 2026PolicyWatcher review · 15:00 UTC

Source

06Review due

High priorityIn force

Supervision and enforcement of AI literacy measures begins

Source: EU AI Office
Category: AI governance
Region: European Union
Review: Reviewed during EU AI Act monitoring

Supervision from 3 Aug 2026Enforcement started · verified 17 Aug

Source

07Review due

High priorityIn force

Full Commission enforcement begins for GPAI obligations

On 2 August the Commission moved from the collaborative phase to full enforcement of general-purpose AI model provider obligations, including additional measures for models with systemic risk.

Source: EU AI Office
Category: Regulatory updates
Region: European Union
Review: Reviewed during EU AI Act monitoring

Enforcement from 2 Aug 2026Fully enforceable · verified 17 Aug

Source

08Review due

High priorityGuidance

Commission publishes Article 50 AI transparency guidelines

Published 20 July and updated 24 July 2026; the guidance covers provider and deployer transparency obligations that apply from 2 August 2026.

Source: EU AI Office
Category: AI governance
Region: European Union
Review: Reviewed during EU AI Act monitoring

02 Aug 2026Obligations apply · 09:00 UTC review

Source

09Review due

High priorityIn force

AI Act Article 50 transparency obligations are now applicable

Source: EU AI Office
Category: AI governance
Region: European Union
Review: Reviewed during EU AI Act monitoring

In force 2 Aug 2026Applicable · verified 17 Aug

Source

Contribution and update model

Registry and operational monitoring

The observatory starts from a curated public census. Catalogued sources become operational only after human review of scope, authority, methodology and evidence role; the calendar therefore uses evidence-ready sources only.

Roadmap queue

Explore

4 active monitoring categories

The Atlas places the Observatory beside the evidence console, quality and trust surfaces, methodology and roadmap.

Open the site Atlas