Rules on Prohibited AI Practices Have Been Introduced Since February

The rules on prohibited artificial intelligence (AI) methods have been in force since February under the European regulation (AI Act), explains lawyer Sara Rocha, TMC associate at CMS Portugal, in an interview with Luca.

The European regulation, published in the Official Journal of the European Union (EU) on 12 July, enters into force 20 days after publication and must be implemented by August 2026.

AI “is the topic of the decade and as such is not regulated”, and with its publication it applies directly to all Member States, says TMC Technology, Media and Communications Officer at CMS Portugal.

“The regulation provides for extended periods of time for the entry into force of some of its measures. Following its publication, the regulation will enter into a transitional period, with most of its rules applying for 24 months from the date of entry into force, with different application periods for individual issues,” he points out.

Of particular note is “the case of the rules on prohibited artificial intelligence methods, which will apply for six months,” that is, from February, he adds.

“The main reason for adopting these implementation stages is that, given the complexity of some of the measures, we want to give companies time to adapt,” explains Sarah Rocha.

The OV Act “is based on a risk-based system that specifies certain uses of AI that are considered unacceptable and therefore prohibited, such as subconscious manipulation of behavior, exploitation of people considered vulnerable, social assessment by government or private parties that leads to discrimination, and remote biometric identification in public spaces in real time by law enforcement, with some exceptions,” he explains.

Asked about the expectations for its implementation, the lawyer stresses that “its main objective is to promote the development and use of AI in the EU and to guarantee a high level of protection, while trying to stimulate the use of AI and the development of the sector”, and also provides for “specific obligations to protect citizens”.

Where once “the debate was about the pros and cons of cloning or using technology, the debate now oscillates between the benefits of using AI – relieving us of some work, helping to combat some diseases, speeding up many response measures, etc. The automation of processes – and the inconvenience or fear that comes with using them – increases social inequality and discrimination by turning into uncontrollable systems or making it impossible to distinguish between a human and a machine.”

In response, the European legislator approved a regulation containing “68 definitions, 113 articles, 13 annexes and 180 summaries”, which is “a diploma with a very complex structure and reading intended only for a niche that is already familiar with this type of language, most of these rules are not easy to understand, and excessive complexity is not always the best option for innovation, which is one of the main features.”

In this context, the lawyer says that it is expected that it will be possible to “continue to develop artificial intelligence systems that are useful to the population, without losing speed in their development, but ensuring that these tools respect citizens and the legal system in which they operate.” They are included, but are aware of these difficulties.”