Which AI use cases in government and public sector are high risk under the EU AI Act?

Annex III point 5(d): AI systems intended to evaluate and classify emergency calls or to dispatch or set priority for emergency first response services (police, fire, medical aid) are high risk. Pure transcription that performs a narrow procedural or preparatory task may fall outside it under the Article 6(3) exceptions, but alerts that influence triage are in scope. An AI agent that speaks with callers directly, for example on a non emergency line, must also tell them they are interacting with AI (Article 50).

Annex III point 5(a): AI systems used by or on behalf of public authorities to evaluate the eligibility of natural persons for essential public assistance benefits and services, or to grant, reduce, revoke or reclaim them. A fundamental rights impact assessment (Article 27) is required before a public body deploys it. A design that scores people over time on their social behavior or personal characteristics and leads to unrelated or disproportionate detrimental treatment would fall under the Article 5(1)(c) prohibition on social scoring.

Annex III point 8(a) makes AI high risk when it is intended to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts. Classifying a filing by legal theme and flagging it for selection or combined resolution goes to which law applies and how, not only to organizing the file, so the narrow procedural task exception in Article 6(3) is unlikely to apply here, unlike a tool that only summarizes or indexes a file for a person to read.

Annex III point 4(a) lists AI systems intended to be used for the recruitment or selection of natural persons, in particular to place targeted job advertisements, to analyze and filter job applications and to evaluate candidates. Screening, ranking and scoring applications is therefore high risk. A component limited to a narrow procedural task, such as booking interview slots or answering process questions, can fall outside the high risk category under Article 6(3), but only if it does not materially influence the outcome and does not profile people, and that assessment must be documented (Article 6(4)). Deployers of the high risk part must follow the instructions for use, assign competent human oversight, keep logs, inform workers' representatives and inform candidates that a high risk system is used (Article 26). An organization that builds its own screening system becomes its provider, with conformity assessment duties. The chatbot part also carries the Article 50 disclosure duty.

Annex III point 5(a): systems used by or on behalf of a public authority to evaluate a natural person's eligibility for essential public assistance benefits and services, or to grant, reduce, revoke or reclaim them. Recital 58 names "social services providing protection" among those essential services, and both tools sit inside exactly that: whether a family receives a child protective investigation and the services that can follow. Article 6(3) lets an Annex III system avoid the high risk tier when it only performs a narrow procedural task, improves a completed human decision, or is a preparatory step, but that exemption does not apply where the system performs profiling of natural persons, which this model does from personal and administrative data. So profiling closes the exemption; it is not a separate route into the high risk tier on its own.

Annex III point 3(b): AI systems intended to be used to evaluate learning outcomes in educational and vocational training institutions at all levels are high risk. Scoring that determines access to an institution or the level of education a student will receive is also covered by points 3(a) and 3(c). Schools and exam bodies that use such a system have the deployer obligations of Article 26.

Assistants that talk with residents must tell people they are interacting with AI (Article 50(1)), and AI generated text published to inform the public on matters of public interest must be disclosed unless it has had human review under editorial responsibility (Article 50(4)). Internal translation that neither talks with people nor is published carries no specific obligation. Translation can also sit inside an Annex III process, such as examining asylum, visa or residence permit applications (point 7(c)) or evaluating emergency calls and dispatching emergency services (point 5(d)). Whether the translation component is itself high risk depends on its intended purpose (Article 6(3) exempts systems that only perform a narrow procedural task); either way it should be governed with that high risk process.

Annex III point 8(a) makes AI high risk when it is intended to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts. Tools for prosecutors fall under point 6(c) if they evaluate the reliability of evidence, and tools that assist the examination of asylum, visa or residence applications fall under point 7(c). Under Article 6(3) a system that only performs a narrow procedural task or a preparatory task, such as organizing a file or transcribing and summarizing it for the person who decides, may not be high risk, but the provider must document that assessment (Article 6(4)). Summaries of internal legal advice for government lawyers, as Amsterdam plans, are generally outside Annex III.

Permit and license decisions are not listed as such in Annex III, so officer decision support is usually minimal risk, and an assistant that talks to applicants carries the Article 50 transparency duty. The exceptions are permits in an Annex III area: examining applications for visas and residence permits (point 7) and evaluating eligibility for essential public assistance benefits and services (point 5(a)) are high risk. Solely automated decisions with legal or similarly significant effects on a person fall under GDPR Article 22 whatever the tier.

Prioritizing inspections of businesses and premises is not a use listed in Annex III, so such a system is usually not high risk. The assessment changes when it scores natural persons, such as individual licensed professionals or sole traders, and the inspectorate acts as a law enforcement authority: assessing the risk that a person offends, or profiling persons in the detection or investigation of criminal offenses, is high risk under Annex III point 6 (d) and (e), and predicting that a person will commit a criminal offense based solely on profiling is prohibited by Article 5(1)(d). GDPR applies wherever sole traders, home based businesses or named professionals are scored.

Contract review and sourcing are not among the Annex III high risk uses, so an internal assistant that makes no decisions about natural persons is minimal risk (with the Article 4 AI literacy duty). If a negotiation bot chats directly with supplier staff, Article 50(1) applies and it must tell them they are dealing with an AI system, unless that is obvious from the context. Public authorities using AI in procurement should still check national public procurement rules on transparency and equal treatment of bidders.

Annex III point 7(c) makes AI high risk when it assists public authorities in examining applications for asylum, visas or residence permits, including assessing the reliability of evidence. Applicant facing information assistants that give general guidance fall under the Article 50 transparency duties (limited risk). Evidence classification, routing and interview support used in the examination are likely high risk, unless the provider documents under Article 6(3) that a component only performs a narrow procedural or preparatory task. That exception never applies to a system that profiles natural persons, which matters for routing on personal attributes or risk profiles.

Annex III point 5(a) makes AI high risk when it is used by or on behalf of public authorities to evaluate the eligibility of natural persons for essential public assistance benefits and services, or to grant, reduce, revoke or reclaim them. An assistant that only explains rules and guides applications carries the Article 50 transparency duties (limited risk); one that screens or scores eligibility falls under point 5(a), and a public body deploying it must carry out a fundamental rights impact assessment first (Article 27).

Not listed by name in Annex III. Recital 59 says AI systems used by tax and customs authorities in administrative proceedings should not be classified as high risk law enforcement systems; this covers the revenue, misclassification and valuation side of the risk targeting on this page. Contraband and narcotics targeting that leads to seizures sits nearer criminal enforcement: Annex III point 6 (law enforcement) can apply where customs acts in criminal enforcement and the system profiles a natural person, for example building a case around a seizure such as the heroin and methamphetamine finds Indian Customs describes. Annex III point 7(b) covers risk assessment of natural persons entering the Union; it does not apply here, since this use case scores goods, consignments and trading entities, not natural persons. Where a natural person is profiled, GDPR's profiling rules apply; Article 22 covers a decision based solely on automated processing that produces a legal or similarly significant effect, which this use case's human in the loop design (an officer decides every flagged case) is intended to keep out of scope on its own, a separate GDPR question from AI Act classification. Criminal customs investigations fall under the Law Enforcement Directive (EU) 2016/680 instead of the GDPR.

Risk selection for administrative tax audits is not listed in Annex III, and Recital 59 says systems used by tax and customs authorities in administrative proceedings should not be treated as high risk law enforcement systems. Use in criminal tax investigations (Annex III point 6, law enforcement), or evaluating the eligibility of natural persons for public assistance benefits run through the tax system (Annex III point 5(a)), can make it high risk. When individuals are scored in administrative tax work, the GDPR applies, including its profiling rules (Member States may restrict some rights for taxation matters under Article 23). Article 22 applies when a decision with legal or similarly significant effect is taken solely by the model. Criminal investigations fall outside the GDPR and under the Law Enforcement Directive (EU) 2016/680 instead.

Minimal for a system level register of systems and owners with no monitoring of individual employees; it is not listed in Annex III and is the instrument deployers use to meet obligations such as the Article 26 duties for high risk systems and the Article 49 registration of Annex III systems in the EU database. Limited where the plain language assistant that staff and auditors query is not obviously an AI system to its users: under Article 50(1) its provider must then design it so people are told they are dealing with AI. Possibly high risk under Annex III point 4(b) on worker management if the discovery process monitors or evaluates the behavior of individual employees rather than staying at the level of systems and owners.

Classifying documents and extracting data for a person or process to use is usually minimal risk. Even inside an Annex III area, a system that only performs a narrow procedural task, such as splitting and classifying documents, can fall outside the high risk category under Article 6(3); the provider must document that assessment and register the system (Article 6(4) and Article 49(2)). The picture changes when extraction materially influences decisions in Annex III areas, such as eligibility for public assistance benefits (point 5(a)), creditworthiness (point 5(b)) or asylum, visa and residence permit applications (point 7), where the whole system must be assessed as potentially high risk. The Article 6(3) exception never applies when the system performs profiling of natural persons.

The tool drafts documentation for a social worker to check and sign rather than deciding on services or eligibility itself, but a Care Act assessment or similar eligibility document is what an adult social care eligibility decision rests on, which sits close to Annex III 5(a): AI systems used by a public authority to evaluate eligibility for essential public assistance benefits and services. A deployment designed as a preparatory drafting step, feeding into but not replacing that eligibility assessment, can rely on the Article 6(3)(d) derogation for AI performing a preparatory task to an assessment relevant for the purpose of the Annex III use case, rather than the assessment itself. Relying on that derogation carries its own duties, and they fall on the provider: a council that builds its own tool this way is the provider under the Regulation, and it must document why the system is judged non high risk (Article 6(4)) and register itself and the system in the EU database (Article 49(2)). That is different from Article 49(3), which requires a public authority to register its use of a system that actually is high risk, and does not apply once the derogation holds. Separately, the provider of the generative model may owe the Article 50(2) duty to mark the drafted text as AI generated. If a council reused the same transcripts or drafts to feed a scoring or triage model, that downstream use would need its own risk assessment (see the referral risk triage use case).

Research and drafting support for lawyers in firms and companies is not listed in Annex III, so it is normally minimal risk with AI literacy duties. Annex III point 8(a) makes it high risk when a judicial authority, or someone on its behalf, uses AI to research and interpret facts and the law and to apply the law to a concrete set of facts, or when it is used in a similar way in alternative dispute resolution, so a deployment for courts, tribunals or arbitration needs its own classification.

A score that only orders a collector's worklist or chooses the timing of a reminder is not listed in Annex III. The same score becomes high risk under Annex III point 5(b) if it is used to evaluate the creditworthiness of natural persons, and may fall under point 5(a) if a public authority uses it to grant, reduce, revoke or reclaim essential public benefits or services. Keep the scope to prioritization and document where the line sits.

Testing controls over transactions and systems is not an Annex III use. Controls that monitor and evaluate individual employees' behavior, such as trading or access conduct, can fall under Annex III point 4(b), so the design decides the tier.

Generating training content is not listed in Annex III. Providers of tools that generate synthetic audio, image, video or text content must mark the output as AI generated (with an exception for assistive editing that does not substantially alter the source), and deployers must disclose deep fakes, such as an avatar or voice that resembles a real person and would falsely appear authentic (Article 50(2) and (4), with the definition in Article 3(60)). If the same system evaluates learning outcomes or decides access to training that affects a person's work, Annex III point 3 (education and vocational training) and point 4 (employment) must be checked, and those parts can be high risk.

Document review for a party in civil litigation or an internal investigation is not listed in Annex III, so it is usually minimal risk. It becomes high risk where a law enforcement authority uses AI to evaluate the reliability of evidence in the investigation or prosecution of criminal offenses (Annex III point 6(c)), or where a judicial authority uses it to research and interpret facts and law (point 8(a)). Prosecutors and investigators should classify each use against those points.

Transcribing and summarizing meetings for the participants is minimal risk. It becomes high risk under Annex III point 4(b) if transcripts are analyzed to monitor or evaluate individual workers' performance or behavior, and inferring participants' emotions from their voices or faces at work is prohibited by Article 5(1)(f). Recording and transcription also need a lawful basis and clear information to participants under GDPR.

Screening tender documents of companies is not a use listed in Annex III, so such a system is usually not high risk. The tier changes if the system scores named individuals, such as officials or sole traders, to assess the risk that they commit a criminal offense on behalf of a law enforcement authority, which falls under Annex III point 6. GDPR applies wherever named officials, bidders who are individuals or whistleblowers appear in the data.

An automated valuation model values the collateral, not the person, so it is not itself listed in Annex III; the EU Mortgage Credit Directive treats property valuation (Article 19) and the creditworthiness assessment of the borrower (Article 18) as separate steps, and Article 18(3) says the creditworthiness assessment must not be based predominantly on the value of the property exceeding the amount of credit, or on an assumption that the property's value will increase. The valuation becomes relevant to Annex III point 5(b), creditworthiness assessment of natural persons, only where its output is built into a separate system that evaluates the borrower's creditworthiness, and whether that happens depends on how the lender designs the credit decision, not on the valuation model itself.

Answering onboarding questions and orchestrating provisioning is limited risk: under Article 50(1) the assistant must be designed so that employees are told they are interacting with AI, unless that is obvious. It becomes high risk under Annex III point 4(b) if it is used to make decisions on the terms or termination of the work relationship, to allocate tasks based on individual behavior or personal traits, or to monitor and evaluate new hires' performance or behavior, for example to judge probation.

Counting people and estimating flow without identifying or categorizing individuals is not listed in Annex III. The tier changes with the design: remote biometric identification of people in a crowd falls under Annex III point 1(a), biometric categorization under point 1(b), and use by or on behalf of law enforcement to assess or profile individuals under point 6. Real time remote biometric identification in publicly accessible spaces for law enforcement is prohibited under Article 5(1)(h), with narrow listed exceptions. France ran its Paris 2024 trial under a national law that authorized eight listed situations, after an opinion from the data protection authority CNIL, according to the French press.

Claims fraud, waste and abuse detection is not itself listed in Annex III. Annex III point 5(a) only covers AI used by or on behalf of public authorities to evaluate a natural person's eligibility for essential public assistance benefits and services, or to grant, reduce or revoke them, which can cover statutory health schemes run by or for public bodies such as Medicare; it does not cover a private health insurer denying a member's own claim. For a private insurer, only Annex III point 5(c), risk assessment and pricing in life and health insurance, can make a system high risk, and claims fraud, waste and abuse detection by itself is not covered by it. Provider level detection that never decides a patient's own entitlement to care stays outside Annex III either way.

Internal intake, extraction and coding for review by safety staff is not listed in Annex III and is usually minimal risk. A public facing reporting assistant must tell people they are talking to an AI under Article 50. The main obligations come from pharmacovigilance law and good pharmacovigilance practices, which require validated, inspectable processes, and from GDPR rules on health data.

Triage of phishing, endpoint, network and cloud alerts for an organization's own cyber defense is not listed in Annex III. Recital 55 of the AI Act says that components intended to be used solely for cybersecurity purposes should not qualify as safety components, so the agent does not fall under Annex III point 2 (critical infrastructure), and for this scope the tier is minimal. The design changes that when the agent triages identity, data loss prevention, insider risk or user behavior alerts in a way that scores or monitors individual employees: monitoring and evaluating the behavior of persons in a work relationship falls under Annex III point 4(b), so that scope needs its own high risk assessment before it goes live. The Article 50(1) duty to disclose AI interaction does not apply because it is obvious to a reasonably well informed analyst that they are working with an AI agent. An operator that lets AI act autonomously on network or operational technology controls should assess that design separately, and reading employees' emails and sign in data remains subject to data protection law.

Classifying and summarizing text feedback is minimal risk. The tier changes if the system infers emotions from customers' voices or faces in calls or video: emotion recognition based on biometric data is listed as high risk in Annex III point 1(c) and triggers the Article 50(3) duty to inform the people exposed. Analyzing feedback from employees to evaluate individual workers moves it towards Annex III point 4(b), and emotion recognition in the workplace is prohibited by Article 5(1)(f), except for medical or safety reasons.

People must be told they are interacting with an AI system, and synthetic voice or text must be identifiable as such (Article 50). Reminding people of existing bookings and disclosure alone are limited risk. A missed appointment score used by or for a public authority to grant, reduce, revoke or reclaim access to healthcare or other essential public assistance and services, for example deciding who is offered funded transport, can fall within Annex III point 5(a), and profiling of natural persons within Annex III rules out the Article 6(3) exemption. Using the score only to decide who gets extra reminders or support does not by itself place it outside Annex III when that support is itself the assistance being granted.

Annex III point 5(b) excludes AI used to detect financial fraud from the high risk credit scoring category, but a system that in effect decides on creditworthiness is high risk, and remote biometric identification is high risk under point 1(a), which excludes one to one biometric verification. When a public authority uses the model on claims for public benefits, point 5(a) can apply, because it covers AI used to grant, reduce, revoke or reclaim benefits and has no fraud exception. Keep fraud detection separate from the credit or eligibility decision and use biometrics only for one to one verification.

It depends on where the system runs. Classifying and routing a bank's or insurer's correspondence is not a use listed in Annex III, so it is minimal risk: the AI literacy duty of Article 4 applies, and the Article 50 duty to tell people they are dealing with AI does not, because the system does not interact with the sender. Used by or for a public authority in a benefits process covered by Annex III point 5(a), the provider can treat it as not high risk only while it performs a narrow procedural or preparatory task under Article 6(3); the provider must then document that assessment before it goes live (Article 6(4)) and register the system in the EU database (Article 49(2)). If the system evaluates eligibility for benefits or profiles the people who write in, it is high risk, so those judgments stay with people.

Annex III point 4 lists AI used for the recruitment or selection of natural persons (4(a)) and AI used to make decisions affecting promotion, or to allocate tasks based on individual behavior, personal traits or characteristics (4(b)). A marketplace that ranks employees for internal roles or allocates projects on the basis of inferred traits is therefore high risk. Recommending learning content or mentors to an employee who chooses freely is usually not. Deployers of the high risk part must inform workers' representatives and the affected employees before use (Article 26).

Used only for practice and feedback, the simulator is limited risk. Article 50 requires that people know they are interacting with AI unless that is obvious from the context, as it usually is in a training session, and the provider must mark synthetic voice or text output as AI generated in a machine readable format. It becomes high risk under Annex III point 4(b) if its scores are used to evaluate the performance of workers or to decide on their promotion or termination, and can fall under point 3(b) when a vocational training institution uses it to evaluate learning outcomes. Inferring trainees' emotions from voice or face in the workplace is prohibited under Article 5(1)(f), except for medical or safety reasons.

An information assistant must tell people they are interacting with AI (Article 50). It is not high risk as long as it does not evaluate eligibility for public assistance benefits or services (Annex III point 5(a)); an assistant that starts to pre assess eligibility should be reassessed.

A taxpayer assistant must tell people they are interacting with an AI system (Article 50). It is not listed in Annex III as long as it only informs and applies fixed rules. It becomes high risk under Annex III point 5(a) if it evaluates eligibility for, or grants, reduces, revokes or reclaims, public assistance benefits (which can include benefits paid through the tax system). Recital 59 says systems used for administrative proceedings by tax and customs authorities are not high risk law enforcement systems; audit selection and risk scoring are covered on a separate page.

A 311 assistant must disclose that it is AI (Article 50). It is not high risk while it only informs and creates service cases. If it evaluates or classifies emergency calls or sets dispatch priority for police, fire or medical services, it falls under Annex III point 5(d) and becomes high risk.

A tool that only drafts a narrative for a named officer to review, correct and sign is not listed in Annex III: it does not evaluate the reliability of evidence or a person's risk of offending or reoffending, the law enforcement uses Annex III point 6 covers. Article 50(4)'s disclosure duty for AI generated text does not apply either way, because that duty covers only text published to inform the public on matters of public interest, and an internal police report is not published for that purpose; officer review is not what exempts it. Article 50(2) is the provision that does apply: a provider whose system generates synthetic text from a transcript, rather than only lightly editing text the deployer already supplied, must mark its output in a machine readable, detectable format. That places the tier at limited, not minimal. It would move toward Annex III point 6, and a high tier, if the same kind of system were extended to judge witness credibility or predict reoffending rather than to draft a narrative.

An internal tool that monitors and classifies regulatory publications for staff makes no decisions about natural persons, so it is not listed in Annex III and is not a prohibited practice under Article 5. Staff know they are using an AI tool and its summaries are not published to the public, so the Article 50 duties to inform users and to disclose published generated text add little for the deploying organization. Article 50(2) still requires the provider of a system that generates text to mark its output, in a machine readable format, as AI generated: usually the vendor, but an organization that builds its own summarizer can itself be that provider, which is what puts this use case at the limited tier rather than minimal. Beyond this and AI literacy (Article 4), no specific obligations apply. General model risk and third party rules still apply.

An assistant that gives general legal information and explicitly declines to advise on a person's specific case is not listed in Annex III; it carries the Article 50 duty to disclose that the person is interacting with AI. If a deployment were changed to assess a person's eligibility for legal aid itself, rather than only inform and refer, Annex III point 5(a) on essential public assistance services could apply, and the design should be reassessed.

Producing information videos is not listed in Annex III. Providers of tools that generate synthetic video must mark the output as AI generated, and deployers must disclose deep fakes, such as an avatar modeled on a real staff member (Article 50(2) and (4)). A system that also decided who gets a benefit or service would need a separate assessment under Annex III point 5.

A visitor facing assistant must make clear that people are interacting with AI (Article 50), and synthetic speech should be identifiable as AI generated. It is not high risk. It would change if the camera feature were used to identify or categorize visitors by biometric data, which a guide does not need: remote biometric identification, biometric categorization and emotion recognition are high risk under Annex III point 1, and biometric categorization that infers sensitive traits is prohibited under Article 5.

A drafting copilot whose output an agent reviews and sends falls under the transparency tier at most. When replies are sent without human review, customers interact with the AI system directly, and Article 50(1) requires that they are informed unless this is obvious from the context. Article 50(2) separately requires the provider of a system that generates text to mark its output in a machine readable format as artificially generated, whether or not a person reviews the draft. It becomes high risk only if it is used for a purpose listed in Annex III, such as evaluating eligibility for public benefits or creditworthiness (point 5), or evaluating the performance of the agents who use it (point 4).

Article 50(1) requires that people who interact directly with an AI system are informed of it, unless this is obvious from the context, as it usually is for an internal assistant. The system would be high risk only if it were intended for an Annex III purpose, such as assessing the creditworthiness of natural persons (point 5(b)) or making decisions on or evaluating workers (point 4(b)).

Drafting letters for human approval is not listed in Annex III. The decision the letter communicates may come from a separate high risk system, such as credit scoring (Annex III point 5(b)) or a public body's eligibility decision on benefits (point 5(a)); the drafting tool does not make that decision. Article 50(2) requires the provider of an AI system that generates text to mark the output as artificially generated, which puts this on the limited risk (transparency) tier; this includes an organization that builds its own drafting tool. Article 50(2) does not apply where the AI has only an assistive function for standard editing and does not substantially alter the input data or the semantics of the output.

The translation service generates synthetic sign language video, so under Article 50(2) the provider must mark that output as artificially generated in a machine readable format. A deployer must disclose the video as AI generated under Article 50(4) only where the signer is photorealistic enough to count as a deep fake: content that resembles an existing person and would falsely appear authentic. There is no human review or editorial responsibility exemption for video; that Article 50(4) exemption applies only to generated text published in the public interest. Article 50(1), the duty to tell someone they are dealing with AI, applies only if the avatar interacts with a person directly, for example in a conversational digital human channel; a video played on a screen or on request does not by itself trigger that duty.

An internal drafting assistant that an official reviews is not listed in Annex III. Article 50(4) requires disclosure of AI generated text published to inform the public on matters of public interest, unless it has undergone human review and a person holds editorial responsibility, which this design provides. If the tool is used to evaluate eligibility for public assistance benefits or services rather than to draft, Annex III point 5(a) can apply.

Organizing and summarizing consultation responses for analysts does not decide on individuals and is not listed in Annex III, so no high risk obligations apply. If AI generated text is published to inform the public on matters of public interest without human review and editorial responsibility, Article 50(4) requires disclosure.

Tools that support staff in searching, deduplicating and proposing redactions are not listed in Annex III (point 5(a) covers eligibility for public assistance benefits and services, not access to documents), and every release decision stays with an officer. A public facing request assistant that talks to requesters would carry the Article 50(1) transparency duty.

Drafting internal policy text for human approval is not an Annex III use and has no direct effect on individuals. The Article 4 AI literacy measures still apply to the staff who use it.

Finding buildings and pools in images and checking them against a register is not a use listed in Annex III, and Recital 59 says systems used by tax and customs authorities in administrative proceedings should not be treated as high risk law enforcement systems. The tier rests on that scope, not on the human check. GDPR still applies to owner records and to the tax data used in the match, and the tier would change if the system were extended to rank natural persons by fraud risk.

Assessing organizations as vendors is not an Annex III use. If assessments score individual natural persons, such as sole traders, check the design against Annex III and data protection rules. The EU AI Act also shapes what to ask AI vendors, since providers of high risk systems carry specific obligations.

Processing public text and broadcasts to find health events is not a use listed in Annex III and makes no decision about a person. GDPR still applies where social media posts or other personal data are processed, and the assessment would change if the system were used to profile individuals.

Analyzing complaints in aggregate to find causes is not listed in Annex III, is not a practice prohibited by Article 5 and does not decide on individuals. It does not interact with the public, so the disclosure duty in Article 50(1) does not apply; the machine readable marking of generated text in Article 50(2) is a duty of the provider of the generative model or system that writes the summaries. If the same system decided individual complaint outcomes or redress, or its themes were used to evaluate the performance of individual complaint handlers (Annex III point 4), that design would need its own assessment.

A research and drafting aid that a policy official or legal drafter reviews before use is not listed in Annex III. Article 50(4) exempts AI generated text that has undergone human review and carries a named person's editorial responsibility from the disclosure duty that otherwise applies to AI generated content published on matters of public interest, which is how the UK and New Zealand tools on this page are designed to work. If a tool moved from drafting support to deciding the outcome of a request for a public service or benefit, Annex III point 5(a) could apply instead.

Drafting internal or public help content that a person reviews and publishes is not a prohibited practice under Article 5 and is not listed in Annex III, so it is minimal risk. The articles are not a direct AI interaction, and the Article 50(4) disclosure for AI generated text published to inform the public does not apply where a person reviews the text and holds editorial responsibility. The Article 50 transparency duties do apply to chatbots that later answer customers from the articles.

Classifying the organization's own purchase lines into categories is not listed in Annex III and is used internally by procurement staff, so no specific obligations apply beyond AI literacy. The data can still contain personal data, for example in purchasing card and expense lines, which brings GDPR duties. Using the classified card and expense lines to monitor or evaluate individual employees would move the system towards Annex III point 4 (employment and worker management) and a high risk assessment.

An internal assistant that drafts requirements artifacts and test cases for engineers is not listed in Annex III and does not interact with the public, so no specific obligations apply beyond AI literacy (Article 4). The system under test may itself fall under the Act.

Matching entries between internal financial records is not a use listed in Annex III and is not a practice prohibited by Article 5. Operators knowingly use an internal AI tool, so no Article 50(1) disclosure is needed. If a generative model drafts the explanations or journals, the provider of that system may have to mark its output as AI generated under Article 50(2). The AI literacy duty of Article 4 applies to the bank as deployer.

Processing supplier invoices is not an Annex III use case, is not a practice prohibited by Article 5 and does not involve decisions about natural persons, so it is minimal risk and the AI literacy duty of Article 4 applies. Approvers who ask questions in chat use an internal tool they know is AI; if that is not obvious to the people using it, the provider must also inform them that they are interacting with an AI system (Article 50(1)).

Summarizing public threat reporting for an organization's own cyber defense is not an Annex III use, and recital 55 of the AI Act says components used solely for cybersecurity purposes do not qualify as safety components. The tier stays minimal while the assistant works on public sources and an analyst reviews and takes responsibility for what is published. It changes if the same tooling is used to profile or score employees.

An internal drafting and analysis assistant for auditors that makes no decisions about natural persons. It would need reassessment if used to evaluate individual employees' behavior or performance, which falls under Annex III point 4(b).