Best Practice Guide: Responsible Use of Generative AI in Advertising
The Advertising Association has released a voluntary Best Practice Guide for using Generative AI responsibly in advertising, developed under the Government and Industry-led Online Advertising Taskforce. It turns high-level principles into practical steps, so teams can move fast without risking trust, compliance, or brand equity.
Built with input from an expert working group and the Advertising Standards Authority (ASA), the guide operationalises the ISBA/IPA principles from 2023 into day-to-day workflows. It's a flexible framework that fits alongside UK GDPR, the Equality Act, and the ASA's co- and self-regulatory system.
What's inside the guide
The guide focuses on eight principles with clear actions for advertisers, agencies, and media owners. Here's the headline view with immediate steps you can put to work.
- Transparency: Label AI-generated or AI-edited content where relevant, disclose AI use when it's material, and keep prompt/output logs. Use content provenance signals where available.
- Data use: Only use data you're entitled to use. Minimise PII, set retention limits, and verify vendor data policies and training data licensing.
- Fairness: Test outputs for bias across audience segments. Use diverse datasets and add human review for sensitive categories.
- Human oversight: Keep a human in the loop for briefing, approval, and escalation-especially for high-risk or regulated claims.
- Harm prevention: Apply safety filters, fact-check claims, and run "red team" checks for misinformation, stereotyping, or unsafe content.
- Brand safety: Set clear guardrails (topics, tone, sources). Vet vendors, use allow/block lists, and add deepfake detection where relevant.
- Environmental considerations: Prefer efficient models, batch non-urgent tasks, and monitor compute usage to reduce energy impact.
- Continuous monitoring: Track issues, keep model/change logs, and review performance, bias, and safety at regular intervals.
SME-friendly version
A streamlined SME edition focuses on the principles with the highest impact for small teams-transparency, data use, human oversight, and brand safety. It's lighter to implement without losing the essentials.
Why this matters for marketers
This isn't theory. It's a workable set of rules that helps you cut risk, keep ads compliant, and protect consumer trust-while moving faster on creative and media operations.
Rt Hon Ian Murray MP, Minister for the Creative Industries: "I am pleased to see the publication of this timely industry-led guide, which is a key output of the Online Advertising Taskforce's AI Working Group. This work supports the Government's ambitions to ensure advertising remains trusted and makes the most of the opportunities AI can offer, helping the sector innovate responsibly."
Stephen Woodford, CEO, Advertising Association: "This new guide is designed to support responsible adoption of AI in advertising to ensure that the work of our industry can continue to be trusted by the public. In the words of the ASA, all advertising must be 'legal, decent, honest and truthful' and it must remain so as our industry embraces AI and the many benefits it can bring."
Mark Lund OBE, Deputy Chair of the Online Advertising Taskforce: "Developed collaboratively by a sub-group of the Taskforce containing advertisers, agencies, media and tech companies, this accessible guide provides practical tools to help practitioners meet the opportunities and challenges of Generative AI, whilst upholding the standards that are the foundations of consumer trust."
Immediate actions for your next quarter
- Appoint an AI lead and define RACI for creative, media, legal, and data teams.
- Audit where GenAI is already used across creative, targeting, personalisation, and analytics.
- Create an SOP for prompts, approvals, content labelling, and record-keeping.
- Run a data rights check: consent, licensing, retention, and vendor terms.
- Set up bias and claims testing for high-reach and sensitive campaigns.
- Lock in brand safety rules: topics, tone, sources, and high-risk exclusions.
- Pick efficient models and track compute to reduce environmental impact.
- Train teams on ASA rules, AI disclosures, and incident response.
Governance and resources
The guide complements existing UK law and advertising codes. For reference, see the ASA's system of ad regulation and UK GDPR guidance from the ICO.
If you want practical training, templates, and workflows that fit these principles, explore AI for Marketing.
Download the guides
Access the materials via the Advertising Association:
- Full Guide
- SME Version
- Key Principles at a Glance
Ongoing feedback and updates
These principles are voluntary. The Advertising Association welcomes feedback and will review the guidance regularly to reflect changes in technology, regulation, and industry practice.
The goal is simple: use GenAI responsibly so your brand stays trusted, your campaigns stay effective, and your team moves with confidence.
Your membership also unlocks:
