The EU AI Act has arrived: how Unilever is preparing

The EU AI Act^[a] is the world’s first comprehensive AI law aimed at mitigating the risks of AI use and is widely considered to become the blueprint for other similar AI regulatory regimes developing around the world.

But at Unilever, responsible use of AI has been a priority for years, as Andy Hill, Chief Data Officer, and Christine Lee, Chief Privacy Officer, explain.

End-to-end assurance: the smart way to work with AI

Unilever’s AI assurance journey began in 2019 when, as the global debate on digital transparency and privacy began to escalate, the business started to review its practices on data and AI ethics.

In 2021 Unilever strengthened its business principles to commit to responsible, ethical and fair use of data. Over the following years, we started building our AI assurance process, scaling this up for the generative AI boom. It has now become part of the way we do business.

“Regulatory compliance is a key component of our Responsible AI Framework, and we are proactively monitoring and addressing upcoming legal developments that may impact Unilever,” explains Christine Lee, Chief Privacy Officer.

To ensure regulatory compliance, potential new projects using AI systems at Unilever are assessed by a cross-functional team of subject matter experts, including our partners at Holistic AI. They review the needs of the project, managing risks and suggesting improvements or mitigation strategies that might be needed prior to deployment, as well as any ongoing monitoring.

This triage system is further supported by Unilever’s Responsible AI Principles, which reaffirm our commitment to developing, deploying and using AI technologies in line with Unilever’s Code of Business Principles, legal and regulatory requirements, and UN standards.

Robust risk management

This process is now fully integrated across the organisation, with the team recently hitting their 150 ‘projects assured’ milestone.

“Taking proof of concept projects using AI systems through a thorough assurance process at an early stage is enabling us to be more innovative and fully deploy trustworthy AI systems more quickly,” says Andy Hill, Chief Data Officer.

With over 500 AI systems in operation globally, ranging from AI-driven R&D that enables faster innovation cycles to machine-activated stock control and generative AI-powered consumer experiences in the marketing space, having a Responsible AI Framework that governs the development, deployment and usage of AI is business critical for Unilever.

“We see potential in the use of AI to drive productivity, creativity and growth at Unilever,” adds Andy.

“With augmentation, we’re supporting our teams through learning, enhanced decision-making and new experiences. And by creating autonomous systems, we believe AI can drive productivity within our business.

“As the deployments of these systems grow, we cannot underestimate the importance of making sure it works in a responsible manner.”

Keeping pace in a fast-moving sector

Having full visibility of Unilever’s AI estate ensures we are able to effectively manage risk and more easily keep pace with new global legal frameworks such as the EU AI Act and the US White House Executive Order on AI.

“Although Unilever has developed legal and ethical guardrails for AI, risks around issues such as IP rights, data privacy, transparency, confidentiality and AI bias can remain, as legal frameworks can lag behind the rapidly evolving technology,” explains Christine.

Ongoing vigilance and scalable processes that can adapt to developing regulations is the most effective way to ensure compliance in this fast-moving landscape, and this is where Unilever’s AI assurance system has the advantage.

“As demand for AI systems, technology and capabilities continues to grow, we expect regulations to evolve as different countries discuss their own approaches to governing AI,” says Andy.

“We will continue to ensure Unilever stays in step with legal developments that affect our business and brands – from copyright ownership in AI-generated materials to data privacy laws and advertising regulations.

“I’m very excited to see where the integration of digital innovation and the latest technologies can take us, but we also need to make sure that our data is well governed and that whenever Unilever uses AI, it is in an effective and responsible manner,” he adds.