Future Proof The Authority Stack
Independent Cross-Jurisdictional Publication Global AI Liability Desk 24 April 2026
Agent Liability Global Desk
Agent Liability Global Desk

Editorial Standards

These standards govern every piece of content published on this site. They are published openly so that readers, cited sources, and potential partners can verify the basis on which we operate. Version 1.0, dated 24 April 2026.

1. Independence and the editorial firewall

No carrier, regulator, vendor, law firm, consultancy, or other third party pays for placement, ranking, framing, prominence, or inclusion in any content on this site. This applies to articles, briefings, jurisdiction trackers, framework comparisons, tools, and any other form of publication.

Editorial decisions are made by editorial staff. Commercial relationships (sponsorships, research commissions, partner listings) are handled by a separate function and have no influence over what is covered, how it is framed, or whether a party is included. The two functions operate on separate budgets and separate decision chains. A commercial relationship with any entity does not guarantee coverage of that entity, and the absence of a commercial relationship does not exclude coverage.

Future Proof Intelligence is the publisher of this site. FP is also building its own AI agent certification methodology, published at agentcertified.eu/methodology-v2.html. The certification arm and the editorial arm operate independently. Coverage of the certification methodology on this site is subject to the same evidentiary and citation standards as coverage of any other methodology. FP's commercial interest in the certification programme does not grant it preferential editorial treatment on any publication in this network.

2. How we select what to cover

We cover matters that carry genuine regulatory, market, or operator-impact significance for organisations deploying autonomous AI agents across jurisdictions. The primary criteria are:

  • A statute, regulation, or supervisory guidance has entered or is about to enter the legal record in any of the jurisdictions we track and creates obligations for operators or providers of AI systems.
  • A court decision, enforcement action, or supervisory opinion establishes or clarifies a principle that affects the allocation of liability for AI agent conduct.
  • A carrier, syndicate, or insurance programme announces a named product that directly addresses AI agent liability exposure, with verified capacity and geographic scope.
  • An industry body, standards organisation, or government authority publishes a framework that is relevant to compliance, certification, or underwriting practice in the AI agent context.
  • A cross-jurisdictional development creates regulatory divergence or convergence that is material for operators deploying agents across multiple markets.

We do not cover press releases, promotional announcements, or vendor claims that have not been verified against primary sources. We do not cover content whose primary purpose is to generate traffic to a product or service rather than to inform a reader about a substantive development.

3. How we cite

We cite primary sources wherever they exist. The standards for each source type are as follows.

EU legislation. Cited by official designation and article number. Example: Regulation (EU) 2024/1689 of the European Parliament and of the Council (the AI Act), Article 26(1). Where a recital is cited, the recital number is given.

National legislation (non-EU). Cited by the official short title used in the enacting jurisdiction, with section or article number. For US federal legislation, the United States Code citation is given. For state legislation, the state code section is cited. For UK legislation, the official short title and section number as published on legislation.gov.uk are used. Where the legislation is in a language other than English, we note the official title in the original language and provide the basis for any translation we use.

Court decisions. Cited by full case name, tribunal or court, and date of decision. Where a case number or neutral citation is available, it is included. Quotations from judgments are given in full, with paragraph or page reference.

Supervisory guidance and opinions. Cited by the issuing authority, the official title of the document, and the publication date.

Standards (ISO, NIST, BSI, etc.). Cited by the official standard designation and the clause or section relevant to the claim made.

Trade publications and research reports. Named with the publication or organisation name, the title of the relevant article or report, and the publication date. We do not treat trade publication claims as verified fact.

Carrier products. Cited by the official product name as stated in the carrier's launch announcement or product documentation, together with the date of the announcement.

4. How we handle errors

We correct errors. The classification and handling of corrections follows the distinction between material and non-material errors.

A material error is one that, if left uncorrected, would cause a reader to hold a false belief about a fact of regulatory, legal, or market significance. This includes incorrect article numbers, incorrect dates of legal effect, misattributed statements, and incorrect descriptions of carrier product terms. Material errors are corrected with a dated correction note appended at the top of the affected page, stating what was originally published and what the correct position is. The original text is not deleted; it is struck through so that the change is visible.

A non-material error is one that does not affect the substantive accuracy of the content. Non-material errors are corrected in place, with a footnote noting the correction date.

A log of all material corrections is maintained at /corrections.html. Readers who believe they have identified an error are invited to write to [email protected].

5. Conflicts of interest

Future Proof Intelligence is the publisher of this site and of the four sister publications in the Authority Stack (agentliability.eu, agentcertified.eu, agentinsured.eu, insureyouragent.com). FP has a commercial interest in the development of the AI agent insurance and certification market, as it expects to offer certification services and, in time, to derive revenue from the partner programme described in the Partners page.

These interests are disclosed here and on the Partners page. They do not affect the editorial independence commitments stated in Section 1. Content about FP's own methodology is held to the same evidentiary standards as content about any other methodology. Where FP's certification work is referenced in an article, the connection to the publisher is noted.

Editorial staff are required to disclose any personal financial interest in an entity covered in content they produce. Such interests are reviewed before publication.

6. Right of reply

Any organisation or individual named in content published on this site has a right of reply. A reply request should be sent to [email protected] with the subject line "Right of Reply" and a reference to the specific content. We will acknowledge receipt within five working days.

Replies that are factually substantive and relevant to the content in question will be incorporated as updates to the relevant page or appended as clearly labelled response notes. We reserve editorial judgement over the form of inclusion.

7. Future commercial activity

In 2026, this site operates on an editorial-only basis. No paid placements, sponsored rankings, or commercial partner listings appear on any page. All inclusion is editorial and free of charge.

We anticipate introducing a paid Partners tier in 2027 for verified institutional participants, including carriers, brokers, consultancies, and implementation partners that meet published inclusion criteria. The criteria for paid partnership will be published openly before the programme launches. Paid partnership will not affect editorial coverage decisions, carrier rankings, or the framing of any analysis on this site. All paid placements will be visibly labelled as commercial. The structure and pricing of the 2027 programme are set out in outline on the Partners page.

8. Versioning

These standards are version 1.0, dated 24 April 2026. They will be reviewed annually. Proposed revisions will be published for a minimum 30-day public comment period before adoption.

Verification across the network

The same editorial standards apply across all five publications in the Future Proof Authority Stack. Readers can verify consistency by reviewing the editorial-standards page on each sister site.