We Built Diageo’s First Marketing AI Agent

Learn how Voxly started helping Diageo scale responsible marketing compliance with AI agents back in Feb 2024. 

Will AI compromise Responsible Marketing? 

AI has reshaped media and advertising. We see it in our LinkedIn feeds, and in the posters and newspaper ads we glance at every day on the train ride home after work. 

But does this mean a loss of control and an unintentional move to lower quality? (“AI Slop”)

Or does it mean that marketers can now be more creative and move faster? (“AI nirvana”)

And what happens to purpose and values? Of course brands will want speed, better KPIs and lower costs from AI, but is it worth the risk if used incorrectly? 

Let's take the example of the highly regulated beverage-alcohol industry. Here, marketing communications have always been done in a socially responsible way. This is essential for protecting consumers and to ensure that communications remain compliant with regulatory standards. AI won’t change this. In fact, when implemented correctly, AI can drive ethical communications by ensuring compliance standards are met more efficiently and consistently, without compromising protection for consumers and brands.

Diageo wanted a better way to manage Responsible Marketing

This case study focuses on Diageo - a global leader in the beverage-alcohol industry - who operate brands like Johnny Walker, Baileys, Don Julio, and Guinness across 150 countries. 

Their commitment to responsible marketing is governed by the Diageo Marketing Code - and they are widely recognized as leading the way in communications that are ethical, legal, and socially responsible.

However, responsible marketing is getting harder in the age of AI. Diageo generates tens of thousands of brand and advertising assets annually. Enforcing the Diageo Marketing Code relies on manual reviews by brand managers and compliance experts. Every piece of copy, static image, and video storyboard has to be scrutinized against multi-layered local and global regulations. 

The primary challenges in maintaining high standards include:

• The explosion of AI-generated content can outpace manual compliance checks, creating a bottleneck that risks letting errors slip through. 

• Complexity in Enforcement: Responsible marketing rules are nuanced and open to interpretation, making them difficult to enforce consistently across diverse brands and global markets.

• Legal and Reputational Risk: Failure to catch non-compliant assets can lead to severe consequences, including fines, advertising bans, and lasting damage to brand reputation.

• Slower Time to Market: With advertising review and sign-off consuming up to 50% of the creative process, slow approval cycles lead to missed market opportunities and reduced agility.

• Inefficient Use of Resources: Manual, repetitive checks can be a poor use of human capital and also lead to duplicated effort and high training costs across teams and agencies.

  
  

What Was Built: An AI Agent for Responsible Marketing

Diageo partnered with Voxly to design and build an AI agent called Dominic. It was specifically trained against the different sections of the Diageo Marketing Code and rolled out as a Proof of Concept. It worked by scanning image, text, and video assets and returned structured compliance recommendations - flagging potential issues, referencing relevant sections of the DMC, and giving reviewers an actionable starting point to think through creative challenges in the context of responsible marketing. 

Voxly and Diageo worked together iteratively to test the agent's performance and accuracy across a range of content types, brands, and markets. Rules were refined to reflect the nuances of specific geographies - the United States was among the priority markets - and adapted to the particular sensitivities of individual brands. For example, a Johnnie Walker campaign carries different contextual considerations than a Baileys activation.

From a user standpoint, the AI Agent was designed to fit in smoothly with the way people already worked.

Voxly connected the agent via API into Diageo's existing Digital Asset Management (DAM) platform, supplied by Sitecore, so that compliance review could become a native part of the asset approval workflow rather than a separate step. A Microsoft Teams extension and a conversational chatbot were also developed, which would have allowed users to access compliance recommendations within their day-to-day working environment.

The Proof of Concept was rolled out internally to users for testing and validation. The focus was on accuracy, building trust, and demonstrating that AI-generated compliance recommendations could meet the standard required in a regulated business environment.

Results: Faster Better More Responsible Advertising

The project delivered measurable outcomes across every key performance indicator.

Review time per asset fell from five to ten minutes down to approximately sixty seconds — a reduction of around 90%. Compliance sign-off, which had previously taken three or more days, was compressed to approximately one hour. 

The AI agent also demonstrated over 90% accuracy against DMC expert ratings during testing - all while being engineered to handle up to 100,000 assets per year - a capacity that reflects the real operational demands of a business at Diageo's scale. Based on the results, a full company-wide rollout was estimated to generate in excess of £500,000 GBP in annual savings.

The compliance results spoke for themselves, but we also greatly valued the learning - understanding how to scope, test, and govern an AI system within a regulated marketing workflow. That knowledge is now shaping how we think about AI adoption across brand and marketing within Diageo”.

Dana Radu, Diageo. Marketing Risk, Governance and Product Lead.

These numbers matter, but the more significant outcome may be strategic: Diageo now has direct, practical experience of what it takes to implement an agentic AI system inside brand and marketing. That knowledge - about data preparation, model evaluation, human-in-the-loop design, and change management - also greatly benefitted Diageo.

Strategic Implications: What Marketing Leaders in Regulated Industries Should Take Away

This case study is about more than just about using AI to save time on compliance checks. It's also about the factors to consider when using AI in a process that carries genuine risk.

Start with a standard. The Diageo Marketing Code was the foundation on which the AI agent was built. Without a clear, written compliance framework, there is nothing to train against and no way to evaluate accuracy. Organisations that want to automate compliance review need to invest first in codifying what good compliance looks like — not just as a legal document, but as a set of evaluable criteria.

Accuracy must be tested not assumed. Voxly and Diageo ran months of iterative testing. For any AI application in a regulated context, a structured evaluation programme - including market-specific and brand-specific use cases - is a must.

Integration reduces resistance. Embedding the AI agent into Sitecore and Microsoft Teams, as well as a standalone tool, met a specific design challenge. The friction of switching between platforms is often what causes adoption of new tools to fail. If the AI capability lives inside the tools teams already use, uptake is faster and the behavioural change required is smaller.

Human oversight is a feature not a bug. The AI agent generated compliance recommendations, not compliance decisions. 

The system was designed to augment the judgement of trained reviewers - giving them a faster, more consistent starting point - not to replace it. Organisations should be cautious about any AI compliance tool that removes human accountability from the loop.

A Proof of Concept is a genuine learning vehicle. Diageo used the PoC not just to validate a product, but to develop institutional understanding of agentic AI. The experience of scoping, building, testing, and deploying - even at limited scale - produces knowledge about what these systems can and cannot do. 

---

Summary of KPIs & Milestones

• The first purpose built Marketing AI Agent at Diageo (PoC was delivered in Feb 2024). 

• Individual asset reviews cut from 5-10 min to 60 sec per asset.

• Compliance sign-off from 3+ days to 1 hour.

• Over 90% tool accuracy vs. Diageo DMC experts. 

• £500K+ in estimated annual savings, had the service gone fully live across the company.

---