Starting August 2, Article 50 of the EU Artificial Intelligence Act comes into force, requiring companies to tell people when they are interacting with an AI system rather than a human.
The rule follows growing frustration among French consumers. A study conducted by Flashs for Hostinger found that more than three quarters of French respondents, 77 percent, believe a company should always disclose when an AI assistant is responding to them. When including those who only want disclosure for sensitive topics, that figure rises to 92 percent, according to the study.
What the law requires
The legislation states that providers must ensure AI systems designed to interact directly with people are built so that those people are informed they are dealing with an AI system.
Article 50 includes an exception: the disclosure obligation does not apply when it is obvious to a reasonably well informed, observant and circumspect person, given the circumstances and context, that they are interacting with a machine, such as an automated messaging service.
The article also exempts AI systems legally authorized to detect, prevent, investigate or prosecute criminal offenses, subject to safeguards for the rights and freedoms of others, unless those systems are made available to the public to report a crime. That exemption does not apply to chatbots or phone conversations with an adviser.
A question of trust
Lea Paolacci, head of the news division at Flashs, said French consumers remain strongly attached to knowing who they are speaking with and to being able to reach a human adviser when they feel the need. She said they want to keep control over the relationship.
Paolacci said that beyond the performance of these tools, user trust is also at stake.
