AI Use Policy

In short.

• Every article carries an AI level from 0 to 5, shown next to the byline, or on The Fintech Times in a line at the foot of the piece, so you can see how much of it AI made.
• A named human editor signs off on every published word. No article on any of our titles is published at Level 5 (fully autonomous), and none will be without a published change to this policy.
• The levels are honest about heavy AI use. Most of our news is drafted by AI and edited by a person (Level 2). We say so on the article rather than in a footnote.
• AI-generated images carry a visible watermark naming the model that made them.
• AI never sources its own material. When AI drafts or analyses, an editor chose the source.
• We do not consent to our content being used to train AI models without written agreement. See our Copyright & Takedown Policy.
• This policy is how we meet, and go beyond, the transparency obligations of the EU Artificial Intelligence Act (Article 50).

1. Why this policy exists

Disrupts Media's five publications are B2B trade press read by professionals who depend on us for accurate, sceptical reporting. AI tools are now part of how our journalism is made, and a large part of it. This policy sets out plainly where and how we use AI, where we do not, and how we tell you, so you can judge our work with full information.

Three principles sit behind everything below:

• A human editor signs off on every published word. No content is published without editorial review and explicit sign-off by a named editor.
• Disclosure is proportionate and visible. Every article shows its AI level. Routine tools such as spellcheck need no disclosure; everything beyond that is on the scale.
• Readers come first. When AI shaped a piece, you see that before you read it, not buried in a footer nobody finds.

In this policy, "we", "us" and "our" mean Disrupts Media Limited, the publisher of the Sites.

2. The AI level: six levels from 0 to 5

Every article on every Site carries one AI level. The level describes how much of the published text AI produced and how much a person reviewed before publication. It is set when the piece is produced and confirmed by the editor who signs it off.

How to read the number. Levels 0 and 1 are human-written. Levels 2 and 3 are AI-drafted and human-reviewed, and the difference between them is how much the editor changed. Levels 4 and 5 are where a person stops reading every word before it is published. We define 4 and 5 so you can see where we are on the scale, not because we are there.

Contributed and supplied content. Contributors who hold an author profile on one of our Sites self-certify the AI level of every article they supply, on this scale, and we keep that certification on record alongside our own internal tracking. A contributed piece carries the level its author certified, or a higher level if our own editing added AI use; never a lower one. A contributor who intends to supply a piece at Level 4 or 5 must agree its handling with our editorial team first. Those records are internal and are not published.

For press releases and news stories we receive from companies and agencies, we cannot yet confirm how much AI was involved in producing the material sent to us; the level on the resulting article describes our own production from that source. We are working on this, and expect provider text watermarks to give us a way to do it during the first quarter of 2027.

Images are marked separately (§3). Quotations are never AI-generated at any level.

What stays the same at every level from 1 to 4:

• We do not invent facts, statistics, named people or quotes.
• We do not publish opinion pieces drafted by AI.
• AI never sources its own material. When AI drafts or analyses, an editor pointed it at a specific trusted source: a press release, a dataset we hold, a research paper, an industry report, a regulatory filing. AI does not run open-ended searches, scrape unverified material or choose sources of its own initiative. The editor is responsible for what AI is asked to look at.
• A named human editor reviews and approves before publication.

2.1 How the level appears on an article

• Next to the byline: a small marker reading "AI level 2 of 5" (or the relevant number), styled as disclosure rather than decoration, so the number is visible before you start reading. Hovering over or tapping the marker shows the one-line statement behind it, for example: "AI level 2 of 5: drafted by our AI editorial assistant from the company's announcement; fact-checked, edited and signed off by [Editor Name]." The marker links to this policy.
• On The Fintech Times, for now, at the foot of the piece instead: the same one-line statement as a closing line, linking to this policy. The Fintech Times is moving to the same publishing platform as our other titles and will carry the byline marker when it does.
• In our RSS feeds: each item carries its level as a category, so feed readers can filter on it.
• In the page code: the level is present as machine-readable metadata.

Level 0 is shown as well. A missing marker is a fault, not a statement.

2.2 The basic tools that need no disclosure

Spellcheckers, grammar tools, autocomplete, formatting helpers and search are treated as ordinary editorial infrastructure, the same way we do not disclose use of a dictionary. A piece produced with only these tools is Level 0.

3. Images

What it covers: any image displayed alongside an article that was generated by an AI image model.

Disclosure: every AI-generated image carries a visible watermark in the bottom-left corner of the image, naming the model that produced it, for example "Created by Gemini AI" or "Created by Imagen AI". Our automated image pipeline names the generating model in the watermark on every image it creates, across all five publications.

What this does not cover: editorial photography, supplied press imagery, stock photography and infographics carry no AI watermark.

Image marking is independent of the article's AI level. A Level 0 article can carry an AI-generated hero image, and it will say so on the image.

4. The editor sign-off rule, and the Level 5 gate

This is the bright line behind the whole policy.

• No content is published without editor review and explicit sign-off by a named editor. This applies to every piece on every Site, at every level from 0 to 4.
• No Site publishes at Level 5. Our systems are built so that an article cannot be saved at Level 4 or Level 5 unless that path has been deliberately enabled for a named title and article type, and no such path is enabled.
• Moving any title or article type to Level 4 or Level 5 requires a published change to this policy first, with the change recorded in the document history below, before the first such piece appears. We will not do it quietly.

Editorial responsibility. Disrupts Media Limited holds editorial responsibility for everything published on the Sites, and the editor named on each piece holds it for that piece. Contact details are in §8. This is the arrangement under which we rely on the editorial-control provision of Article 50(4) of the EU AI Act for AI-drafted text (§7), and the AI level is how we go further than that provision requires.

If you read a piece on a Disrupts Media publication, a named editor put their name to it. That commitment does not change as AI tools improve; the level tells you how much of the work the tools did.

5. What this means for readers

• On every article, on every Site: the number from 0 to 5 next to the byline, with the one-line statement behind it, or on The Fintech Times the statement as a closing line.
• On AI-drafted pieces (Levels 2 and 3): the name of the editor who checked and signed off.
• On AI-generated images: "Created by [Model] AI" watermarked bottom-left.
• On everything: a named human editor signed off before publication.

Disclosure language and placement are consistent across all five publications.

6. AI training and our content

We do not consent to our editorial content being used to train artificial intelligence models or any other automated system that produces derivative output, except where we have agreed that use in writing.

If you operate an AI training crawler or scraper, you may not use our content. This applies whether or not you respect robots.txt and whether or not our content is technically accessible to you.

Our position on copyright and on takedown requests is in the Copyright & Takedown Policy.

7. The EU AI Act and provider watermarks

The EU Artificial Intelligence Act (Regulation (EU) 2024/1689) applies transparency obligations to AI-generated content from 2 August 2026 (Article 50). For AI-generated text published to inform the public, the Act does not require a disclosure where the text has undergone human review or editorial control and a natural or legal person holds editorial responsibility for it. Our sign-off rule (§4) is that arrangement, and this policy is where we document and publish it.

We publish the AI level anyway, on every article, because we think readers should not have to rely on a legal exemption to know how their news was made.

Separately, the AI model providers we use are applying imperceptible statistical watermarks to the text their models generate, to meet their own obligations under the Act. Those watermarks identify text as machine-generated; they carry no information about us or about you. Our AI level is the human-readable disclosure; the provider watermark is theirs, and the two are consistent by design.

8. Publication, access and contact

• This policy is published at /editorial/ai-policy on each of the five brand sites and on the corporate site. The level marker on every article links to it.
• A downloadable PDF version is available for partners, advertisers, regulators and any reader who wants to keep a copy.

For any question about how we use AI in our journalism:

Editorial questions and corrections: corrections@disruptsmedia.com Licensing or use of our content: legal@disruptsmedia.com Post: Disrupts Media Limited, 41 Luke Street, London EC2A 4DP, United Kingdom

9. About Disrupts Media

Disrupts Media Limited Registered in England & Wales, Company No. 09447878 Registered office: 41 Luke Street, London EC2A 4DP, United Kingdom

10. Document history