AI Transparency and Content Disclosure

This page explains how AutomateSEO uses artificial intelligence, what that means for content you publish, and what the law requires of each of us. It forms part of the Terms of Service.

1. What is AI-generated here

Essentially all of it. Unless you have written or edited it yourself, assume it came from an AI model:

The AI models we use are listed on our Subprocessors page. We may substitute an equivalent or better model at any time.

2. What AI gets wrong

You need to know this before you publish anything.

• Facts. Models state incorrect things with complete confidence. Dates, figures, prices, product specifications, legal thresholds and statistics are all commonly wrong.
• Citations. Models can invent sources, or cite a real source that does not support the claim. We ground research in retrieved material to reduce this, but it still happens. Check every citation you rely on.
• Currency. Models have a training cutoff. Anything time-sensitive may be out of date even where the article reads as current.
• Law and regulation. Output may describe rules that do not apply in your jurisdiction, or that have changed.
• Bias. Models reflect biases in their training data, which can surface in how groups of people are described.
• Similarity. Output is not unique to you. A similar prompt from another customer can produce similar text.
• Rights. Generated images and video may resemble existing works or real people. Copyright in purely machine-generated material is uncertain in many jurisdictions, including the UK and US.

None of this makes the product defective — it is the nature of the technology. It does mean the review step is not optional.

3. Your responsibility before publishing

Before anything generated here goes live under your name, you should:

1. Verify the facts, especially numbers, dates, prices, claims about products, and anything a reader might act on;
2. Check the citations actually exist and support what they are attached to;
3. Confirm it is legal in your market — advertising rules, sector regulation, consumer law;
4. Have a qualified person review it if you are in a regulated sector (health, finance, legal, insurance, gambling, or anything requiring professional qualification);
5. Confirm you have consent for any real person's likeness or voice used in video;
6. Decide whether to disclose that it was AI-assisted — see section 5.

If you use automated publishing without per-item approval, you have chosen to accept these risks on every item published. See section 5 of the Acceptable Use Policy.

4. Our role and yours under the EU AI Act

The EU AI Act's transparency rules (Article 50) apply from 2 August 2026. They allocate different duties to different people.

4.1 We are the provider

AutomateSEO is an AI system placed on the market under our own name, so Article 50(2) applies to us: outputs must be marked in a machine-readable format and be detectable as artificially generated.

What we do today:

• generated images, audio and video carry the provenance metadata applied by the underlying model providers, where those providers apply it;
• generated content is identifiable as such within the platform, and every article record is flagged as AI-generated in your account;
• this page is our standing transparency notice.

What is still in progress: a consistent, platform-wide machine-readable marking scheme across every output type, including synthetic text. The AI Act gives generative AI systems already on the market before 2 August 2026 until 2 December 2026 to meet the machine-readable marking requirement, and we are working to that date. We will update this section as the technical standards and the Commission's Code of Practice settle.

4.2 You are the deployer

Article 50(4) applies to you if you publish AI-generated text to inform the public on matters of public interest. If you do, you must disclose that the content is artificially generated.

There is an important exemption. That duty does not apply where the content has undergone human review or editorial control and a natural or legal person holds editorial responsibility for the publication. The review has to be substantive — a genuine editorial check, not a rubber stamp.

In practice, for most AutomateSEO customers:

• Ordinary commercial content — service pages, product guides, how-tos for your own business — is usually not "informing the public on matters of public interest", so Article 50(4) does not bite.
• News, current affairs, health, politics, public safety, or similar topics published to a general audience probably are matters of public interest. If you publish that kind of content, either disclose it or make sure a named person genuinely reviews and takes editorial responsibility for it.
• If you publish without reviewing — for example fully automated publishing — you cannot rely on the editorial exemption. Disclose.

Deepfakes. If you publish AI-generated or manipulated video of a real person, Article 50(4) requires you to disclose that it is artificially generated, and there is no editorial-review exemption for this. Our avatar and spokesperson video features produce exactly this kind of content. Label it.

This is not legal advice. If you publish into the EU at scale, take advice on your own obligations. Penalties under Article 50 reach EUR 15 million or 3% of worldwide annual turnover.

5. Disclosing AI use to your readers

Beyond the AI Act, disclosure may be required or advisable because of:

• Search engine policy. Search engines do not penalise AI content as such — they penalise content produced primarily to manipulate rankings. Being useful matters more than being human-written. Some platforms require AI labelling in their own terms.
• Social platform rules. Meta, LinkedIn, YouTube and TikTok each have AI content labelling requirements for synthetic media. These apply when you publish through our integrations.
• Advertising standards. In the UK, the CAP Code requires marketing communications not to mislead. A synthetic testimonial or a synthetic "customer" presented as real is misleading regardless of any AI labelling.
• Your audience's trust, which is usually the real reason to do it.

We do not add a disclosure to your content automatically, because the right wording depends on your jurisdiction, sector and platform. You can add a standing disclosure to your content templates in Settings.

6. Human oversight built into the platform

• Every generated article can be reviewed and edited before publishing.
• Approval steps can be required for article generation, cluster approval, link enhancement and publishing.
• Automated publishing is off by default and must be deliberately configured.
• All AI agent actions are logged in your account.
• You can stop or roll back a scheduled publication before it goes out.

We recommend keeping approval enabled wherever an incorrect publication would harm you.

7. Data and model training

Your content is not used to train our models or our providers' models. We use commercial API tiers that exclude customer data from training by default. See our Privacy Policy for detail.

8. Questions

AI transparency and compliance questions: legal@automateseo.app