As new EU rules on AI come into force, what do the public think of LLM-generated content?

After being passed in 2024, the EU’s AI Act came into force on Monday, 3 August. It will introduce new legislation around risk assessment, documentation, and compliance, among other things, but it will also include laws governing AI-generated content. Penalties for violating the act could run as high as €35m – or 7% of turnover – for organisations that violate the new laws.

So, what concerns do consumers have about how brands use this content? YouGov Surveys data explores the anxieties the UK public have towards AI-generated material, and how using this material might impact brand trust.

When consumers are asked if it is important or not important that content explicitly states that it was created by AI, 89% say that it is.

The data also shows that seven in ten (70%) adults in the UK have concerns that “there may not be enough rules or oversight for how AI is used.” While the Act will theoretically address this concern, brand trust could still be at risk as companies explore uses for AI-generated content.

We asked Britons about the situations where AI-generated content would negatively affect trust in a brand.  According to our findings, the most commonly selected option was if the content feels misleading or deceptive (70%), followed by content that does not disclose AI usage (68%). Over half (56%) would lose trust if AI replaces humans entirely, while 49% express the same for usage across sensitive topics such as news, health, or politics. One in ten indicate they are not sure which situations would reduce brand trust.

Still, the public are not uniformly worried or gloomy. A subset of the UK say that generative AI can help with generating ideas when people are stuck (41%), and content creation more generally (38%). Will these benefits outweigh the concerns? It may depend as much on how brands respond to these concerns as it does on any regulatory interventions.