The Spanish Agency for Data Protection (AEPD) has announced the start of "preliminary investigation actions to the American company OpenAI", the company that owns ChatGPT, "for a possible breach of the regulations". This service, a generative artificial intelligence, is capable of giving answers to complex topics or holding conversations with a user and has recently begun to reach the general public. The AEPD will also be part of the working group created by the European Data Protection Board (EDPB) to cooperate in the different investigations on this matter.

The Spanish organization, which has acted ex officio, had previously addressed the EDPB to introduce the famous chat at its next plenary meeting – from which this working group has left – "considering that global treatments that can have a significant impact on the rights of people require harmonized and coordinated actions at European level in application of the General Data Protection Regulation", as explained in a statement.

"The AEPD understands that global processing operations that can have a significant impact on people's rights require coordinated decisions at European level," an AEPD spokesperson told Reuters at the time.

The technology has been in the crosshairs of European regulators for some time. At the end of March, Italy blocked access to the service for illegal collection of personal data, which the firm uses to 'train' its artificial intelligence. It also accused OpenAI of failing to prevent minors from accessing the tool. Shortly after, Ulrich Kelber, German data protection commissioner, recalled in statements to the German media Handelsblatt that the country could take similar measures.

Artificial intelligence is also under the magnifying glass of US regulators, although in this case it is not for privacy, but for the too ambiguous or grandiloquent promises of marketing departments regarding this technology. Michael Atleson, a lawyer for the Federal Trade Commission (FTC), warned that a "fashionable" term such as AI can be used lightly and too freely, with exaggerated claims. "You don't need a machine to predict what the FTC might do if those claims are baseless," Atleson said.

The working group was born from a meeting in which EDPB members discussed the measures taken by Italy. The idea is that it serves to "promote cooperation and exchange information on the actions carried out by the data protection authorities," according to the institution itself. The decision of the AEPD to act ex officio is a parallel path "within the framework of its powers and competences as a national supervisory and control authority".

The Agency, in any case, recalls that it advocates "the development and implementation of innovative technologies such as artificial intelligence" But considers that it must be done "from full respect for current legislation." In this case, it considers that "only from that starting point can a technological development compatible with the rights and freedoms of people be carried out".

The OpenAI chat, of which a new version recently arrived, is not connected to the Internet and does not work as a search engine, but generates answers based on probability: the system decides which word is most likely to make sense based on the previous ones you have written and also based on what you have been asked for. Thus, it uses its interactions with them to 'learn' and improve its algorithm. This was what triggered the alarms of the Italian agency, which acted for "the lack of information to users and all interested parties whose data is collected by OpenAI" and for "the absence of a legal basis that justifies the collection and mass storage of personal data, with the purpose of training the algorithms that underlie the operation of the platform. "

  • Artificial intelligence

According to The Trust Project criteria

Learn more