How we disclose, mark and stand behind AI-assisted work.

1. What “AI-assisted” means here

We use generative and analytical AI tools to support specific, bounded tasks, including:

• Research synthesis, competitor analysis and keyword/topic discovery.
• First-draft copy, content outlines, FAQs and captions.
• Data analysis, reporting summaries and roadmap suggestions.
• Image and asset concepting where applicable.

In every case, a qualified person reviews, edits and approves the output. The strategy and the final decision are human.

2. How we mark AI-generated content

When content we deliver or publish on your behalf is materially AI-generated (for example, AI-produced images, synthetic audio/video, or substantially AI-written text), we apply transparency markings appropriate to the medium and the law of the markets you operate in:

Visible disclosure

• Images, audio and video that are AI-generated carry a clear, human-readable label where the format and platform allow it.
• Articles and pages that are substantially AI-authored include an editorial note indicating AI assistance and human review.

Machine-readable marking (watermarking & metadata)

• Where supported, AI-generated media is marked in a machine-readable way — through embedded metadata and, where the generating tool supports it, durable watermarking such as C2PA Content Credentials or provider-native watermarks.
• We preserve provenance metadata rather than stripping it, so downstream platforms and detection tools can recognize the content’s origin.

3. Why we do this — the regulatory backdrop

• The EU AI Act transparency duties (Article 50), which require that synthetic image, audio and video content be marked in a machine-readable format and detectable as artificially generated.
• US federal and state-level synthetic-content and disclosure rules, including emerging requirements to label AI-generated media, particularly in advertising contexts.
• Platform policies (search engines, social networks and ad networks) that increasingly require or reward labeled, provenance-tagged AI content.

Requirements differ by jurisdiction, medium and use case, and continue to evolve. This page describes our standard approach; specific engagements may adjust markings to match the client’s regulated context.

4. What we do not do

• We do not pass off unreviewed AI output as finished human work.
• We do not strip provenance metadata or remove watermarks from AI-generated assets.
• We do not generate deceptive synthetic media (impersonation, fabricated endorsements, misleading “real” imagery).
• We do not guarantee that AI content will evade detection — our goal is honest disclosure, not concealment.

5. Your control as a client

You decide the disclosure posture appropriate for your brand and regulatory environment. Before publishing, we agree on labeling conventions, and you approve content prior to release. If your industry imposes stricter marking duties, we adapt the workflow accordingly.

6. Questions

For questions about how a specific deliverable was produced or marked, contact [email protected]. See also our AI Privacy Policy & Data Flow page.