The European Parliament is working on important laws to regulate the use of artificial intelligence. Under this legislation, there may be measures that force chatbots to declare that they are not human.

There are currently several legislative proposals that together could become part of the new European law on artificial intelligence. This law will consist of a set of regulations on the use of AI.

One of the proposals likely to pass legislation is a requirement for chatbots to inform users that they are not real people. That way, chatbots can’t accidentally trick users into thinking they’re talking to a person.

Use of Training Data

Another part of the European Parliament’s AI legislation may require developers of AI products to indicate whether the model was trained on the basis of proprietary information.

Many AI models still use training data collected without permission from the creators. Many AI chatbots also look for information on the internet. To prevent AI from infringing copyrights, there should be a measure whereby copyright owners are paid for their work when used by an AI.

Several lawsuits are currently pending against AI art generators such as Stability AI and Midourney, which are accused of using copyrighted works without permission. Getty Images has also filed a lawsuit against Stability AI alleging that the AI ​​used 12 million copyrighted images in its training set.

Brando Benefei, an Italian politician and MP, says he wants even more transparency on the matter, SiliconANGLE reports: “We want a list of public disclosure of the material used to train it because authors follow different laws and paths may try to get paid for what is being used without their permission”.

No agreement yet

A copy of the draft law, reviewed by Euractiv, shows the legislation may include a ban on the use of AI for predictive policing and emotion recognition. An exception applies to the latter if it is a medical examination for which express permission has been granted.

The talks on Europe’s AI law follow a temporary ban on the use of ChatGPT by regulators in Italy over privacy concerns. Last week, the Italian regulators drafted a list of rules to allow the OpenAI service again.

It’s still too early to say exactly what will be in the legislation for AI, but the end of the talks is in sight. MEPs can reach an agreement ahead of a key committee vote on April 26.