Greenwashing risks associated with artificial intelligence

• technology companies' general unwillingness to share information about their AI models;
• a focus only on the energy use associated with training those models; and
• the ever-evolving nature and growing complexity of AI systems (and the corresponding demands they place on data centre power consumption) such that information on their energy use quickly becomes outdated. 

This lack of clarity raises the question of whether the environmental claims – 'green claims' – made by businesses involved in the creation or the use of AI models are (or are capable of) complying with the UK's rules around greenwashing. 

What is greenwashing?

In broad terms, 'greenwashing' is where a business engages in misleading commercial behaviour which creates an impression that its products or services are more environmentally friendly than they really are.

Greenwashing is a stated priority enforcement area for the UK's consumer law regulator, the Competition and Markets Authority (CMA).  Under newly acquired enforcement powers, the CMA can, amongst other things, fine businesses up to 10% of their global turnover.  With these new powers, which apply to conduct taking place on or after 6 April 2025, the CMA should be expected to start taking a stronger line against misleading claims. 

To-date, the CMA has focused its attention on sectors such as fast-moving consumer goods and fashion, but given AI's very high profile, widespread adoption and high energy use, it is obviously at significant risk of coming under scrutiny.

How is greenwashing regulated?

The Green Claims Code, while not in itself legally binding, sets out the CMA’s view of the law relating to unfair commercial practices and is intended to help businesses understand and comply with their legal obligations in relation not making green claims. 

Those legal obligations are now contained in the Digital Markets, Competition and Consumers Act 2024 (the DMCCA), relevant parts of which took effect on 6 April 2025 (and restating much of the Consumer Protection from Unfair Trading Regulations 2008).  The DMCCA does not refer expressly to greenwashing or green claims: instead, it regulates the making of green claims under its provisions relating to misleading commercial practices.

The Green Claims Code sits alongside two advertising codes administered by the UK Advertising Standards Authority, the CAP Code and BCAP Code (for non-broadcast and broadcast advertising respectively), but these are intended to be consistent with the Green Claims Code.  It is worth, however, looking at the ASA's specific advice on environmental claims under the CAP Code. 

What are green claims?

'Green claims' include claims that suggest or create the impression that a product or service:

• has a positive environmental impact or no impact on the environment;
• is less damaging to the environment than a previous version of the same good or service; or
• is less damaging to the environment than a competing good or service.

A potentially misleading green claim could be a statement that a product or service is environmentally-friendly because it uses renewable energy, but without considering the product's or service's lifecycle, including the manufacturing of the product or, in an AI context, the training of the AI model.

Businesses not only need to take account of the rules around greenwashing when they make explicit environmental claims, they also need to ensure that implicit environmental claims (such as those made through branding, imagery, or trademarks) are compliant.  Further, and perhaps of most relevance to the AI industry given The Economist's findings, businesses need to ensure that any claims made do not omit or hide information in order to give the impression that their services are less harmful or more beneficial to the environment than they really are.

Business-to-consumer environmental claims

For business-to-consumer claims, businesses should consider both whether they are complying with the six core principles listed in the Green Claims Code and whether they can answer 'yes' to each of the statements in the UK CMA's Green Claims Checklist.  While each statement in the Checklist needs to considered in its own right, several are clearly likely to be more relevant in an AI-context, including that:

1. claims are up-to-date and can be substantiated by credible evidence (ideally independently sourced) to show that they are true;
2. claims either relate to the whole lifecycle of a service or relate to one part of a service without misleading people about the other parts or the overall impact on the environment;
3. claims do not miss out or hide information about the environmental impact that people need to make informed choices; and
4. if a comparison is being used, the basis of it is fair and accurate, and is clear for all to understand.

It is likely that many environmental claims made in relation to AI services will be in relation to carbon neutrality or net zero.  Such claims are not, in principle, prohibited, however it is important to ensure that such claims are clear and transparent.  In particular, businesses should provide accurate information about whether (and the degree to which) they are actively reducing the carbon emissions associated with the delivery of their services or whether they are offsetting emissions.  Where they are off-setting, businesses must provide information about the schemes they are using and these schemes should be based on recognised standards and measurements and, importantly, be capable of objective verification.

Business-to-consumer environmental claims

Rather than being consumer focused, many AI services are directed towards other businesses.  While business-to-business environmental claims are not unregulated, the legal framework (established under the Business Protection from Misleading Marketing Regulations 2008) is less stringent: it prohibits advertising that misleads traders and is likely to impact their economic behaviour or injure a competitor. 

While the Green Claims Code has more limited relevance to business-to-business marketing activities that those which are directed towards consumers, meeting its requirements is widely considered to be good practice. 

Businesses using third-party AI services

Where AI is being used, it is frequently provided via the integration of a third-party's supplier's model into a business' existing service offerings.  In such scenarios, where one business supplies services (or products) to another, whether for resale or incorporation, both businesses may be liable for misleading green claims and may have to substantiate them.

In practice, this can require that a business obtains evidence from third parties (such as the underlying AI company) in its supply chain.  This is because all aspects of a service’s environmental impact over its life cycle, including its supply chain, are potentially relevant to the green claims made in relation to it. 

If you would like to discuss any of the issues covered in this article, please contact Aonghus Heatley, a senior environmental and products lawyer in Fieldfisher's Competition, Regulatory and Trade team. Aonghus regularly advises businesses on new and developing legal risks, including greenwashing.