Insights

Browsing AI legislation on laptop.

New AI Rules — What They Mean for You

New rules governing how AI is used in marketing are now in force on both sides of the Atlantic. The short version: this is already handled on the work we do for you, and there’s very little you need to do. Here’s a plain summary so you know where things stand. The specific laws behind it are listed at the foot of this page, for reference. 

What’s changed

Whether your audience is in Europe or the United States, regulators now expect the same two things when AI is involved in marketing:

  • Transparency — people should be told when they’re dealing with AI, and realistic AI-generated images, video or audio should be clearly labelled.
  • Honesty — claims about what AI can do must be truthful, and reviews, testimonials and engagement must be genuine, never fabricated or AI-generated.

The rulebooks differ by region, but the principles are the same everywhere – and it’s those principles which set our standards.

What we’re already doing for you

Much of this simply matches how we already work, since the rules have largely caught up to practices we adopted before they existed. Each of these requirements are simply a reflection of our longstanding practices:

  • Clear “you’re talking to AI” notices on any chatbot or conversational unit we run for you. This has been our standard practice long before any law required it.
  • Honest, substantiated claims. We never – and never will – overstate what our AI tools or your campaigns will do.
  • No fake or AI-generated reviews, testimonials or engagement. We never use fake reviews; we’re all about real people and real engagement, and always will be.
  • Documented human editorial review on AI-assisted copy. Human-in-the-loop reviews have always been a part of finishing any AI-generated outputs.
  • Labelling realistic AI-generated images, video and audio, and preserving the provenance information attached to them. We always have, and always will, value transparency for you and your customers.
  • Vetting the AI tools and vendors we use on your behalf, so disclosure responsibilities are clearly assigned and nothing falls to you by accident. We’re your partner, and we strive to make sure you have all the information you need to do and look your best.
  • Ongoing, role-appropriate AI-literacy training for our team. This is already in place across all Freeman brands, including mdg, and only grows more robust with every update.

What we may need you to check

Only a few things sit outside our scope and are worth a quick look on your side:

  • Confirm any chatbots on your own channels (ones we don’t run) tell users they’re speaking with AI at first contact
  • Keep reviews and testimonials on your channels genuine — no AI-generated or incentivised fakes
  • Preserve the provenance information on assets we deliver — please don’t strip it when re-exporting or re-uploading
  • If you commission or run your own ads featuring realistic AI-generated performers, make sure they carry a clear disclosure
  • If you run your own face-, voice- or other biometric-based campaign technology, make sure you have proper consent in place
  • Let us know if you plan to rebrand or substantially modify an AI tool under your own name, as that can change who holds the legal responsibilities

In short

You don’t need to become an expert in any of this — that’s our job. Wherever your audience is, we’ve built these principles into how we work, and we’ll keep you informed as the rules evolve. If anything here raises a question, just ask your mdg contact.

The rules behind this — for reference

You don’t need to action anything here; these are simply the laws driving the practices above, with links if you’d like to read further.

This summary is for general information, not legal advice. We highly recommend you consult legal counsel if you have any additional questions.