The EU is Bringing Informed Consent to AI

Informed consent has a new frontier, and it may be reaching your social media feed. The EU AI Act went into effect two years ago and was bolstered by new transparency rules as of August 2nd, 2026. The Act is the world’s largest piece of legislation on AI development and deployment, and went into broad effect in 2024. Its initial rules banned certain high-risk applications of AI, such as the public use of facial recognition, and imposed restrictions on lower-risk applications. 

A new rule has gone into effect this month, requiring transparency measures to notify users when they encounter AI-generated content. Any company that has users in Europe now has to provide a watermark or disclaimer with AI outputs, in an effort to fight misinformation.  It also requires companies to disclose when they are using emotion recognition or biometric categorization systems on their users. Because most companies are international, even US companies like OpenAI are required to comply within the EU. 

Transparency Matters Now More than Ever

AI-generated content, from memes to deepfakes, is all over the internet. Because of rapid technological improvement, it is getting harder to discern between real and fabricated content. AI systems are increasingly being used to sway or deceive users, and the EU AI Act’s approach is, in many ways, similar to how we understand consent outside of technology.

While society in decades previously has been comfortable with implied consent in intimate situations, sexual assault survivor advocates have fought to say enthusiastic consent is the bare minimum. How can we know whether someone truly consents to something if there is a risk of coercion or if they were lied to? The same argument can be made for AI.

The EU AI Act gives users the autonomy to decide whether to engage with AI content. When scrolling through social media, users in the EU will now be warned when videos or photos were created with AI, giving them a leg up in determining whether or not they believe what they are seeing. It is a powerful change not only for fighting misinformation, but also for putting the user in the driver’s seat to make decisions for themselves.

What Does this Mean for the US?

Because the EU is a huge user base, it is very possible that companies will comply with the rules globally. This is called the “Brussels Effect.” Coined after the implementation of the  General Data Protection Regulation (GDPR), data protection mandated in the EU was applied outside the EU because it is easier and cheaper than tailoring technology to specific countries. Because tech companies have to update their products to comply with the EU AI Act, it would not be surprising if these disclaimers began appearing in US social media feeds.

Some critics of the EU AI Act's transparency measures argue that they may have an adverse effect on efforts to combat misinformation. If people do not see an AI disclaimer, they may immediately think that the content they are seeing is true, even though humans are just as capable of lying as machines. This critique is especially relevant in the United States, as some companies may choose to comply with EU transparency regulations in the US while others may not. 

In the workplace, discerning AI hallucinations is already proving to be difficult. Often, companies will bring in approved tools like Copilot Chat without providing responsible use education. Because a tool is approved by a company, employees can falsely believe that Copilot’s outputs are completely trustworthy. Just like the EU’s regulations, misinformation is best addressed when tools are transparent, and users are given the resources to understand the risks.

With the new changes to the EU AI Act, informed consent is likely to become even murkier for US users. Given that we may only receive some disclaimers of AI-generated content, we have to be even more cautious when determining whether content is real or fake. We need to keep talking about misinformation and be vocal about the deepfakes we encounter. While we wait for more regulation, we have to empower ourselves.