Reference
Sources register
(formerly Wingman360)
Every claim about a regulation, a standard or a vendor on this site resolves to one of the 47 documents below. Each entry records the document, the revision or edition used, the date it was last checked, and the pages that cite it. Regulation text is quoted verbatim in the guides; nothing here is paraphrased as rule.
Terms used across the guides are defined in the glossary. Guides are re-read against their sources every six months or when a source is revised.
EASA
AMC and GM to Part 21, Issue 2, Amendment 9 (Annex to ED Decision 2019/018/R)
European Union Aviation Safety Agency, 2019. The decision that introduced the certification programme AMC and the means of compliance codes
Cited by Certification-basis compliance matrix: EASA Part 21 and FAA
Applies from 16 October 2025 (Article 8)
Cited by On-premise and air-gapped AI for aviation and defence
As amended, including Implementing Regulations (EU) 2020/639 (point UAS.SPEC.030) and 2021/1166 (point UAS.SPEC.050)
Cited by SORA 2.5 with AI: the ten steps, the UAS operations manual
Applies from 22 February 2026 (Article 16)
Cited by On-premise and air-gapped AI for aviation and defence
As amended, including Regulation (EU) 2020/270 (Part-CAMO) and Regulation (EU) 2021/1963 (Part 145 management system)
As amended, including Regulation (EU) 2020/270 (Part-CAMO, point CAMO.A.300) and Regulation (EU) 2021/1963 (Part 145, points 145.A.70 and 145.A.85)
Cited by AI assistant for EASA Part 145 and CAMO organisations, DOA handbook, MOE and CAME with AI: what EASA requires
Commission Regulation (EU) No 748/2012, Annex I (Part 21)
Points 21.A.15 and 21.A.20, as amended by Commission Delegated Regulation (EU) 2019/897 and Regulation (EU) 2021/699
Subpart J as amended by Commission Delegated Regulation (EU) 2022/201, which introduced the design management system and the current text of points 21.A.243 to 21.A.247
Cited by Certification-basis compliance matrix: EASA Part 21 and FAA, DOA handbook, MOE and CAME with AI: what EASA requires
European Union Aviation Safety Agency, March 2024
Cited by Can AI be trusted for aviation compliance documentation?
EASA Artificial Intelligence Roadmap 2.0
European Union Aviation Safety Agency, May 2023
Cited by Can AI be trusted for aviation compliance documentation?
Consolidated regulations, acceptable means of compliance and guidance material, published in PDF, online and XML formats
Cited by Can AI be trusted for aviation compliance documentation?
European Union Aviation Safety Agency, 10 November 2025. Part (B), proposed DS.AI: DS.AI.010 scope, DS.AI.110 classification, DS.AI.130 risk assessment. Comments closed 10 March 2026
European Union Aviation Safety Agency, 10 November 2025. Part (A) section 2.1.2 on affected stakeholders and Part (B) DS.AI.010 on scope
Cited by Can AI be trusted for aviation compliance documentation?, AI assistant for EASA Part 145 and CAMO organisations, How to choose an AI tool for aviation compliance
European Union Aviation Safety Agency news, 10 November 2025. NPA 2025-07 is the first step of RMT.0742, to be followed by a second NPA in 2026; it covers data-driven AI-based systems, and the framework is to be extended to knowledge-based, hybrid and generative AI systems. Checked 23 September 2026
Checked 23 September 2026 · Cited by How to choose an AI tool for aviation compliance
European Union Aviation Safety Agency. Annex II (Part 145) points 145.A.30, 145.A.45, 145.A.55, 145.A.65, 145.A.70, 145.A.200 with AMC1 145.A.70(a); Annex Vc (Part-CAMO) points CAMO.A.200, CAMO.A.300, CAMO.A.315, CAMO.A.325 with AMC1 CAMO.A.300 and AMC2 CAMO.A.200(a)(6); Annex I (Part-M) point M.A.401
European Union Aviation Safety Agency. Annex II (Part 145) points 145.A.70 and 145.A.85 with AMC1 145.A.70, GM1 145.A.70, AMC1 145.A.70(a), AMC1 and GM1 145.A.85; Annex Vc (Part-CAMO) points CAMO.A.300 and CAMO.A.130 with AMC1 CAMO.A.300 and AMC1 CAMO.A.130
European Union Aviation Safety Agency. Points 145.A.45(e) and (g) and CAMO.A.325 on maintenance data
European Union Aviation Safety Agency. Point 145.A.45 on current maintenance data
European Union Aviation Safety Agency. Point 145.A.45 and AMC1 145.A.45(c), (d) and (g) quoted from the consolidated text, pages 394 to 396; checked 2026-09-20
European Union Aviation Safety Agency. Point CAMO.A.300 and AMC1 CAMO.A.300 quoted from the consolidated text, pages 1126 to 1129; AMC2 CAMO.A.300 read on page 1129; checked 2026-09-20
Cited by AI assistant for EASA Part 145 and CAMO organisations, DOA handbook, MOE and CAME with AI: what EASA requires, On-premise and air-gapped AI for aviation and defence, Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work, How to choose an AI tool for aviation compliance, 145.A.45: maintenance data, CAMO.A.300: the continuing airworthiness management exposition (CAME)
European Union Aviation Safety Agency. Annex II (Part-IS.I.OR) points IS.I.OR.200, IS.I.OR.205, IS.I.OR.210, IS.I.OR.235, IS.I.OR.240, IS.I.OR.250 and their AMC and GM
European Union Aviation Safety Agency. Points IS.I.OR.205 and IS.I.OR.235 on interfaces and contracted activities
European Union Aviation Safety Agency. Point IS.I.OR.205, GM1 IS.I.OR.205 and AMC1 IS.I.OR.205(a) quoted from the consolidated text, pages 108 to 109; checked 2026-09-20
Cited by On-premise and air-gapped AI for aviation and defence, Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work, How to choose an AI tool for aviation compliance, IS.I.OR.205: the information security risk assessment
European Union Aviation Safety Agency. Consolidated Part 21 with AMC 21.A.15(b), Appendix A means of compliance codes, AMC 21.A.15(b)(5), GM 21.A.20 and AMC 21.A.20(c)
European Union Aviation Safety Agency. AMC 21.A.15(b) and Appendix A means of compliance codes, used for the EASA column of the comparison
European Union Aviation Safety Agency. Annex I (Part 21) Subpart J points 21.A.243, 21.A.245 and 21.A.247, with AMC1 and AMC2 21.A.243(a), AMC1, AMC2 and GM1 21.A.243(d), AMC1 and GM1 21.A.247
European Union Aviation Safety Agency. Point 21.A.15, AMC 21.A.15(b) and Appendix A quoted from the consolidated text, pages 111 to 116; checked 2026-09-20
European Union Aviation Safety Agency. Point 21.A.20 and AMC 21.A.20(c) quoted from the consolidated text, pages 131 to 133; checked 2026-09-20
Cited by Certification-basis compliance matrix: EASA Part 21 and FAA, FAA aircraft compliance checklist under 14 CFR Part 21, DOA handbook, MOE and CAME with AI: what EASA requires, 21.A.15(b): the certification programme, 21.A.20: demonstration of compliance and the compliance documents
Easy Access Rules for Large Aeroplanes (CS-25)
European Union Aviation Safety Agency. CS 25.1309 and AMC 25.1309, the EASA equivalent of the FAA advisory circular
Cited by Functional hazard assessment with AI: ARP4761A, AC 25.1309
European Union Aviation Safety Agency, incorporating ED Decision 2025/018/R. AMC1 to Article 11 (SORA 2.5): the route by SAIL to a statement of compliance, a design verification report or a type certificate, and the DVR requirement for M2 or containment claimed at high robustness
European Union Aviation Safety Agency, incorporating ED Decision 2025/018/R. Article 11; AMC1 Article 11 (SORA, source JARUS SORA v2.5, edition September 2025) Sections 0, 1, 3 and 4, Annex A chapters A.1 to A.5 and Annex E; points UAS.SPEC.030 and UAS.SPEC.050 with AMC1 UAS.SPEC.030(2), EASA Form 208
Cited by EASA SC Light-UAS and SC-VTOL: means of compliance with AI, SORA 2.5 with AI: the ten steps, the UAS operations manual
Guidelines on design verification for UAS operated in the specific category, Issue 3
European Union Aviation Safety Agency, 26 September 2023. Sections 2 to 4 and Annex 1 on developing means of compliance; the same page carries the table of means of compliance to SC Light-UAS and the application form
Cited by EASA SC Light-UAS and SC-VTOL: means of compliance with AI
Means of Compliance with Light-UAS.2510, Issue 1, Equipment, Systems and Installation
European Union Aviation Safety Agency, 1 February 2024. Purpose and applicability, including the derivation from OSO #5 and OSO #10/#12 and the exclusions. Read together with MOC Light-UAS.2405-01 and 2410-01 (4 December 2023), MOC Light-UAS.2511-01 (5 May 2022), the functional-test-based MoC for SAIL III and below (26 May 2022) and MOC Light-UAS High Risk.2510-01 for SAIL V and VI (11 February 2025)
Cited by EASA SC Light-UAS and SC-VTOL: means of compliance with AI
European Union Aviation Safety Agency. SC-VTOL-02 Issue 2 of 10 June 2024, VTOL.2000 to VTOL.2010; MOC SC-VTOL Issue 2 (May 2021), MOC-2 Issue 3 (22 December 2022), MOC-3 Issue 2 (2022), MOC-4 Issue 2 final (11 July 2025), MOC-5 Issue 1 proposed (18 July 2025)
Cited by EASA SC Light-UAS and SC-VTOL: means of compliance with AI
European Union Aviation Safety Agency, 17 December 2020 (medium risk, SAIL III and IV) and 22 December 2021 (high risk, SAIL V and VI). Introductory note, applicability, methodology, mitigation means, and the contents list from Subpart A to Subpart H
Cited by EASA SC Light-UAS and SC-VTOL: means of compliance with AI
FAA
14 CFR 145.109, Equipment, materials, and data requirements
Federal Aviation Administration, eCFR. Paragraph (d) on the documents and data a certificated repair station must keep current and accessible. eCFR text as of 15 September 2026, checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work, How to choose an AI tool for aviation compliance
14 CFR 145.209, Repair station manual contents
Federal Aviation Administration, Title 14 Code of Federal Regulations
Cited by AI assistant for EASA Part 145 and CAMO organisations
14 CFR 145.211, Quality control system
Federal Aviation Administration, Title 14 Code of Federal Regulations
Cited by AI assistant for EASA Part 145 and CAMO organisations
14 CFR Part 21, Certification Procedures for Products and Articles (eCFR, current as of 8 May 2026)
Federal Aviation Administration. Sections 21.16, 21.17, 21.19, 21.20, 21.21, 21.31, 21.33, 21.35, 21.41, 21.93, 21.97, 21.101, 21.113 and 21.115
Federal Aviation Administration, via the Electronic Code of Federal Regulations. §§ 21.17 and 21.20 quoted from the current text as of 8 May 2026; checked 2026-09-20
Cited by FAA aircraft compliance checklist under 14 CFR Part 21, 14 CFR 21.20: compliance with applicable requirements
FAA Advisory Circular 25.1309-1B, System Design and Analysis
Federal Aviation Administration, issued 30 August 2024. Failure condition classifications and the probability objectives paired with them
Cited by Functional hazard assessment with AI: ARP4761A, AC 25.1309
FAA Order 8110.4C, Type Certification, Change 7
Federal Aviation Administration, effective 20 October 2023, paragraph 2-3d Certification Plan
Federal Aviation Administration. Certification plan and compliance checklist; referred to, not quoted
Cited by Certification-basis compliance matrix: EASA Part 21 and FAA, 14 CFR 21.20: compliance with applicable requirements
FAA Order 8110.4C, Type Certification, Change 7
Federal Aviation Administration, effective 20 October 2023. Paragraphs 2-3d Certification Plan, 2-4g Issue Paper, 2-4j Certification Basis, 2-5d Completed Certification Plan and 2-6m Compliance Reports
Cited by FAA aircraft compliance checklist under 14 CFR Part 21
Roadmap for Artificial Intelligence Safety Assurance, Version I
Federal Aviation Administration, 2024
Cited by Can AI be trusted for aviation compliance documentation?
JARUS, EUROCAE, RTCA, SAE and other standards bodies
December 2023, technically identical to SAE ARP4761A. Sections 3.2 and 3.4, Appendix A (AFHA) and Appendix C (SFHA). Licensed document, available from EUROCAE
Cited by Functional hazard assessment with AI: ARP4761A, AC 25.1309
EUROCAE ED-279, Generic Functional Hazard Assessment for UAS/RPAS
October 2020, prepared by WG-105. The UAS-level FHA process and its review of ED-135 applicability to unmanned aircraft. Licensed document
Cited by Functional hazard assessment with AI: ARP4761A, AC 25.1309
JARUS SORA v2.5, main body and Annexes A to I
Joint Authorities for Rulemaking on Unmanned Systems, 13 May 2024. The methodology EASA adopted as AMC1 to Article 11; the JARUS text is the source of the ten steps, the phases, the OSO table and the operations manual structure
Cited by SORA 2.5 with AI: the ten steps, the UAS operations manual
NSO public standards database. Status promulgated, 2 April 2019, security class non classified, sponsor Joint Capability Group on Unmanned Aircraft Systems (JCGUAS). The covered standard AEP-4671 Edition B Version 1 is downloadable from the same site
Cited by STANAG 4671 compliance matrix for a military UAS: AEP-4671
SAE International, December 2023
Cited by Functional hazard assessment with AI: ARP4761A, AC 25.1309
Vendor statements and other sources
Anthropic, Is my data used for model training? (commercial products)
Anthropic Privacy Center. Commercial products such as Claude for Work and the API are not used for training by default. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
Anthropic, Is my data used for model training? (consumer products)
Anthropic Privacy Center. Chats are used to improve models only if the user chooses to allow it; incognito chats are never used. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
Claude Developer Platform, Features overview
Lists the platforms through which Claude models are offered: the Claude API, Amazon Bedrock, Google Cloud Vertex AI and Microsoft Foundry. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
EMACC Guidebook, Type Certification Basis Tailoring Guidebook, Edition 1.0
European Defence Agency, MAWA Forum, endorsed 29 January 2014. The description of STANAG 4671, its derivation from CS-23, its applicability limits, and worked example 3.1
Cited by STANAG 4671 compliance matrix for a military UAS: AEP-4671
European Defence Agency, MAWA Forum, approved 4 October 2022, published 12 October 2022. AMC 21.A.15(b), its Appendix with the means of compliance codes, AMC 21.A.15(b)(5), AMC 21.A.20(c) and GM 21.A.20(d)
Cited by STANAG 4671 compliance matrix for a military UAS: AEP-4671
European Defence Agency, Military Airworthiness Authorities (MAWA) Forum, approved 30 March 2021. Points 21.A.15, 21.A.19, 21.A.20 and 21.B.80
Cited by STANAG 4671 compliance matrix for a military UAS: AEP-4671
European Military Airworthiness Certification Criteria (EMACC) Handbook, Edition 3.1
European Defence Agency, MAWA Forum, endorsed 25 September 2018. Sections 1.1 Purpose, 1.2 Applicability, 1.2.1 Tailoring to create the certification basis, 1.4 Information sources, and the STANAG cross-reference table
Cited by STANAG 4671 compliance matrix for a military UAS: AEP-4671
Google Cloud, Gemini is now available anywhere (28 August 2025)
General availability of Gemini on Google Distributed Cloud air-gapped, running in the customer's own data centre. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
Google, Gemini Apps Privacy Hub
How Gemini Apps activity is used to improve services including model training, human review of a subset of chats, and the Keep Activity setting. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
Google, Generative AI in Google Workspace Privacy Hub
Workspace customer data is not used to train models without the customer's permission and is not human reviewed. Checked 17 September 2026
Checked 17 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
Office Open XML File Formats, ECMA-376
The schema each output .docx is validated against in the automated pass
Cited by RDB-1: Regulated Document Benchmark
Data ownership, retention controls, encryption and independent security audit statements for ChatGPT Enterprise and the API. Checked 6 September 2026
Checked 6 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
OpenAI, How your data is used to improve model performance
OpenAI Help Center. Training defaults for consumer plans, business plans and the API, the opt-out, and Temporary Chat. Checked 6 September 2026
Checked 6 September 2026 · Cited by Lavionic Teammate vs ChatGPT, Claude and Gemini for aviation work
RDB-1 working paper v1.0, Lavionic GmbH, September 2026
This page is the published edition of the working paper. Scores dated 6 September 2026
Cited by RDB-1: Regulated Document Benchmark