FREE REGULATORY SCOPE SCREENER · Checklist

AI Regulatory Applicability Screener

Find the regulatory questions to investigate before automating a workflow.

Runs in your browser
No sign-in
Result before any contact step

Illustrative · Example values

Fictional EU and NYC hiring workflow: selected scope questions need review; other laws remain unassessed.

Screen one use case for selected EU AI Act and NYC Local Law 144 review topics. This tool complements risk triage: it identifies scope questions rather than a review-priority score.

Initial coverage is limited to these two rule sets. Privacy, sector rules, other jurisdictions and all other laws are not assessed.

Initial coverage: EU AI Act scope and selected review topics; NYC Local Law 144 employment-tool scope. Other jurisdictions, privacy, sector rules and all other laws are not assessed. This is not a complete obligations or deadline assessment.

Answers stay in this browser tab. No sign-in or contact details are needed. Method reviewed 2026-10-19.

AI and EU scope

1. Is this an AI system?

Confirm the legal definition with the technical owner. A simple rules-based calculation is not automatically AI.

2. Does the use have an EU connection?

Yes if offered or put into service in the EU, deployed by an EU organisation, outputs used in the EU, or other Article 2 connections such as EU import/distribution or affected persons. Headquarters outside the EU is not an exclusion.

3. Might a special role, exclusion or transition provision apply?

For example research, defence, personal non-professional use, open-source conditions, a general-purpose model role, product-specific rules or a previously deployed system. These require a separate review; do not assume an exemption.

EU review topics

4. Could a prohibited-practice category be involved?

Examples include harmful manipulation, certain social scoring, specified biometric practices or emotion inference at work or school. These are examples, not the full Article 5 test.

5. Could the use fall within Article 6 / Annex I or III?

Consider regulated product safety components, biometrics, critical infrastructure, education, employment, essential services, law enforcement, migration or justice. The exact purpose, profiling, exceptions and product rules need review.

6. Could an Article 50 transparency topic be involved?

Direct AI interaction, synthetic content, deepfakes, public-interest text, emotion recognition or biometric categorisation. Duties differ for providers and deployers and have specific exceptions.

NYC hiring and promotion

7. Does the hiring use have a NYC job or employment-agency connection?

Includes jobs at least partly in NYC, remote jobs associated with a NYC office, or use by an employment agency in NYC. Candidate residence alone does not establish job scope; it matters for notice duties.

8. Does it screen or assess specific applicants or promotion candidates?

Any hiring or promotion stage can count. A general resume bank or outreach before someone applies to a specific job is different.

9. Are you the employer or employment agency using it?

A vendor alone is not the employer responsible for the bias-audit duty. If supplying a vendor tool, assess the actual employer/agency use separately.

10. Does the technology produce a simplified employment output using ML, statistical modelling, data analytics or AI?

Examples are a score, classification or recommendation. Confirm the full AEDT and computational-process definitions in the NYC rules with the vendor.

11. Does that output substantially assist or replace discretion?

Yes if relied on alone, weighted more than any other criterion, or used to overrule conclusions from other factors including human judgement. A person making the final decision does not automatically remove scope.

0 of 11 answered

How it works

Answer all eleven questions; choose Not sure when a fact has not been checked. Blank answers cannot produce a result.

Each EU route requires AI status and an EU connection. Topic routes also require the relevant topic. NYC requires all five NYC scope factors. An explicit No closes only that limited route; every Not sure remains in Open facts.

Review indicated means investigate. Needs facts means an unresolved trigger. No trigger identified is limited to the checked route and your answers. None is a clearance or compliance verdict.

Possible exclusions, special roles and transitions require human review. Confirm exact rules, roles, exceptions and dates with a qualified reviewer. The tool does not calculate legal deadlines.

Formula and rules

Routing rules, not legal determinations

EU scope = AI system AND EU connection.
EU topic = AI system AND EU connection AND topic indicated.
NYC scope = NYC nexus AND hiring/promotion candidates AND employer/agency use AND qualifying technology/output AND substantial assistance.
For each AND route: any No = No trigger identified; otherwise any Not sure = Needs facts; otherwise = Review indicated.
A possible EU exclusion never automatically clears the result. All unknown answers remain open.

Assumptions

One actual or planned use case is described, with self-reported facts.
Human final review alone does not exclude a NYC AEDT.
The broad EU purpose and transparency questions are leads for detailed classification, not substitutes for Article 6 or Article 50.

Worked example

Illustrative worked example · example valuesSample

Fictional employer: applicant ranking for an EU operation and jobs associated with a NYC office. A qualifying automated output outweighs every other criterion.

The example answers Yes to AI, EU, Article 6 relevance and all five NYC factors. It reports No for a possible exclusion, prohibited practice and transparency topic. It indicates EU scope/classification and NYC scope reviews; it does not establish a legal classification or permission to deploy.

Evidence to gather

Records that make the result easier to check and act on:

Intended purpose, system definition, accountable owner and provider/deployer roles.
Deployment locations, EU output use, NYC job/agency nexus and first-use dates.
Vendor method, score weighting, human overrides, classification rationale, applicable audit and notice records.
Use the Accountability Toolkit supplement to record reviewer decisions, evidence references, owners and review dates.

What it does not do

Not legal advice, a complete obligations inventory, an audit, certification or permission to deploy.
No system inspection, vendor-document review or independently verified facts.
Rules and dates can change. Re-screen after changes and have a qualified reviewer validate conclusions.

Questions

What does Not sure do?

It keeps the fact open. A route with no explicit No and at least one Not sure shows Needs facts.

Does the tool check deadlines?

No. Application dates, exclusions and transition provisions require reviewer validation against current official sources.

Does this identify every law that applies?

No. Initial coverage is selected EU AI Act and NYC Local Law 144 questions. Privacy, anti-discrimination, sector rules and other jurisdictions are not assessed.

Are my answers sent to Anvaylo?

No. Screening runs in this tab. Download a copy if needed; following the assessment link does not transmit the answers.

Can Anvaylo review the evidence?

You can request a scoped paid assessment. Scope, deliverables and fee are agreed before work starts. It is separate from legal advice, certification and an independent statutory bias audit.