Sample reports that show the output clearly

AI Act sample reports show exactly what structured readiness outputs look like. These examples replicate the format produced by the Evidence Scanner and workspace: executive summaries, obligation mappings, evidence-gap tables, prioritized action plans, and role-specific recommendations. They cover real operational scenarios under current EU AI Act rules — from high-risk classification and Article 50 transparency to GPAI provider duties — without offering legal opinions or guarantees. Use them to understand the shape of deliverables before uploading your own documents.[1][2]

Current law status (May 2026)

• AI literacy measures (Article 4) have applied since 2 February 2025.
• General-purpose AI (GPAI) model obligations (Article 53 and related) have applied since 2 August 2025, including technical documentation, copyright policy, and training-content summaries. Non-EU providers must appoint an authorised representative.[3]
• Transparency obligations (Article 50) for interactive systems, AI-generated content, deepfakes, and certain publications apply from 2 August 2026. Most high-risk AI system rules (Annex III) also start then.
• Proposal note (separate from current law): The Digital Omnibus proposal (November 2025, now in trilogue) would link some high-risk application dates to the availability of harmonised standards and support tools, extend simplifications to small mid-caps, and adjust timelines (up to a maximum of 16 months in certain cases). These changes are not yet in force. Always check the official AI Act Service Desk timeline for the latest status.[4]782651)

Each sample is illustrative only. It demonstrates the structure and depth of outputs from the Evidence Scanner and paid workspace — not a completed compliance file or legal opinion. No certification or guaranteed compliance is implied or provided.

Typical sections in every report include:

• Report card — one-page summary of system classification, applicable legal lanes, and overall readiness score.
• Findings tables — obligations mapped to your described use case, with evidence status and gap analysis.
• Action plan — prioritized, role-aware steps (provider vs deployer) with suggested evidence artifacts and deadlines tied to current application dates.
• Evidence repository view — placeholders showing how uploaded policies, model cards, or training summaries are linked to specific articles.

These outputs help teams translate fragmented legal text into concrete workflow: what to document, who owns it, and how to track updates as new guidance or amendments appear. They reflect current obligations under the AI Act (e.g., Article 4 AI literacy, Article 50 transparency, GPAI rules) while highlighting practical splits between AI system providers, deployers, and GPAI model providers.[2]

Samples are generated from realistic but fictional inputs. Your own reports will adapt automatically to the systems, documents, and roles you describe.

Use this chooser to identify the closest match. Each links to a full downloadable PDF-style example that mirrors workspace output.

Sample report chooser

View the samples

• Recruitment AI readiness report
• Customer-facing chatbot transparency report
• Marketing agency AI literacy and transparency plan
• Non-EU GPAI initial screening report

These excerpts show the exact style and level of detail users receive.

Recruitment AI – Sample Report Card excerpt System name: CV Screening Assistant v2.1 Role: Provider & Deployer (internal HR tool) Classification: High-risk (Annex III, point 4 – access to self-employment and worker management) Applicable now: Article 4 AI literacy; full high-risk obligations from August 2026 (subject to any Omnibus adjustments). Readiness score: 45/100 (strong policy draft, gaps in logging and FRIA)

Findings table excerpt

Action plan excerpt

1. Complete FRIA by Q3 2026 (template provided in workspace).
2. Upload bias audit logs and dataset cards.
3. Assign literacy owner and record completion per official AI literacy Q&A guidance.
4. Schedule quarterly evidence review before August 2026 deadline.

Customer-facing chatbot – Sample Article 50 findings excerpt Key transparency requirement: Inform users they are interacting with an AI system unless obvious (Article 50(1)). Current design: Introductory message present but buried in footer. Recommended label: “This is an AI-powered assistant. Responses are generated and may contain inaccuracies. Please verify important information.”

Marketing agency – Sample action plan excerpt

• Implement machine-readable watermarking or metadata for all AI-generated images and video used in client campaigns (aligns with Code of Practice measures for providers).
• Add visible disclaimers for deepfake-style content or public-interest text.
• Document editorial review process to qualify for Article 50(4) exceptions.
• Track AI literacy training for creative staff (Article 4).

Non-EU GPAI provider – Sample GPAI report card excerpt Model: Foundation LLM v3 (open-weight release) Obligations triggered: Technical documentation, copyright policy, sufficiently detailed training-content summary (template available via EU SEND). Authorised representative: Required before placing on EU market. Systemic risk check: Below current compute threshold but monitor Annex XIII criteria. Next step: Publish summary per official template and prepare downstream information sheet for AI system integrators.[3]

These examples deliberately separate provider and deployer duties, link evidence directly to articles, and flag upcoming dates so teams can sequence work realistically.

• Treating a sample as a universal checklist instead of adapting it to your specific AI system description, role, and context.
• Blurring provider vs deployer responsibilities (e.g., assuming a marketing agency that only deploys GPAI systems has the same technical documentation duties as the original model provider).
• Ignoring phased timelines and preparing only for August 2026 rules while overlooking current Article 4 literacy and GPAI obligations that already apply.
• Uploading generic policies without linking them to concrete system cards or logs, resulting in persistent “evidence gap” flags.
• Expecting the report to replace legal advice or official authority guidance.

• Identify your primary role (provider, deployer, importer, authorised representative) and main legal lane using the table above.
• Review the matching sample report in full via the links.
• Gather existing artifacts: system descriptions, training data summaries, literacy training records, content-labeling procedures, or copyright policies.
• Run a free scan in the Evidence Scanner to generate your own tailored report card and findings table.
• Assign owners and deadlines to the top three prioritized actions, tying them to current application dates.
• Re-scan every quarter or after material changes to your systems, models, or new official guidance.
• Use the workspace to maintain a live evidence repository that maps directly to Article 50, GPAI, or high-risk requirements as they phase in.

Ready to generate your own AI Act readiness report? Start with the free Evidence Scanner to upload your documents and receive a customized report card, findings table, and action plan in minutes. It turns these sample structures into live, trackable outputs for your specific use cases.

Open the Evidence Scanner

FAQ

Are these real legal opinions? No. The samples are illustrative outputs demonstrating the format, depth, and operational focus of reports generated by the Evidence Scanner and workspace. They are not legal advice, official interpretations, or opinions. Refer to primary sources on eur-lex.europa.eu, ai-act-service-desk.ec.europa.eu, and digital-strategy.ec.europa.eu, and consult qualified professionals for legal counsel.

Can I generate something similar from my own docs? Yes. Upload policies, model cards, training summaries, literacy records, or content guidelines. The scanner maps them automatically to relevant obligations (Article 4, Article 50, GPAI Articles 53–55, etc.) and produces customized versions of the report cards, findings tables, and action plans shown here. Outputs update as you add evidence or as the regulatory timeline evolves.

Sources This page draws exclusively from official EU sources including the AI Act Service Desk timeline, Article 113 application dates, AI literacy Q&A, Article 50 transparency FAQ and Code of Practice materials, GPAI guidelines and Q&A, and the consolidated Regulation (EU) 2024/1689. Digital Omnibus proposal status reflects trilogue developments as of April 2026. No law-firm blogs or secondary analyst commentary were used for legal or timeline claims.