The Spanish Agency for Data Protection (AEPD) announced yesterday that it has officially initiated preliminary investigation proceedings against the American company OpenAI, owner of ChatGPT, for a possible breach of data protection regulations.
In this way, Spain joins France and Germany, which are also investigating the use of personal information made by this artificial intelligence. Italy, the United States and China have also announced their intention to legislate to avoid the risks that this new technology entails.
At the same time, last week the AEPD requested the European Data Protection Committee to include the ChatGPT service as a topic to be addressed in its plenary meeting, considering that global treatments that can have a significant impact on the rights of people require harmonized and coordinated actions at the European level in application of the general regulation of Data Protection.
In this way, this committee decided –in the plenary session held yesterday– “to launch a working group to promote cooperation and exchange information on the actions carried out by the data protection authoritiesâ€, regarding ChatGPT.
In a statement, the AEPD explains that this agency “advocates the development and implementation of innovative technologies such as artificial intelligence with full respect for current legislation, since it considers that only from this starting point can it be carried out.” a technological development compatible with the rights and freedoms of the peopleâ€.
But this is not the only front that was opened yesterday to artificial intelligence (AI). The music generated with this technology has become a problem for the industry. Or at least that’s what Universal Music Group (UMG) claims, one of the three largest record labels in the world, which has urged music platforms such as Apple Music or Spotify to block the training of AI models, because they would be violating the rights of author of the songs they use.
UMG considers that certain artificial intelligence systems “could have been trained with protected content” without obtaining the necessary consents and without paying financial compensation to the creators of said works.
“A lot of generative AI is trained on popular music. You could say: I want to write a song with Taylor Swift-like lyrics, Bruno Mars-esque voice, and something like a Harry Styles song. The result you get is because the AI ​​has been trained on the intellectual property of those artistsâ€, explains a source mentioned by the Financial Times.
According to several emails to which the Financial Times has had access, Universal Music Group warns that it will not hesitate to “take steps” to protect the rights of its artists. For this reason, they urge streaming music platforms such as Apple Music or Spotify to block generative AI developers from accessing their music catalogues.
Nor do they rule out taking legal action against any company that does not respect the copyright of artists.
In this sense, Google already has a service called MusicLM, which generates music from text. This model is trained with 280,000 hours of music and that, for the moment, the company has not released due to a “risk of possible misappropriation of creative content.” The results are spectacular, say its developers: “Our experiments show that MusicLM outperforms previous systems both in audio quality and in adherence to text description. It can transform whistled and hummed melodies according to the style described in a text caption.
This technology creates music without the need for an artist involved. You just have to enter a short description of what you want and the AI ​​will do the rest. For example, he has been able to create “a fusion of reggaeton and electronic dance music, with a spacey, otherworldly sound.”
OpenAI, the company behind ChatGPT and Dall-E, has also developed Jukebox, capable of generating music from scratch, simply by entering a genre, artist and lyrics.
It is not the first time that artists have shown concern about the possibilities of generative artificial intelligence. Last January a group of creators sued three companies dedicated to digital art (Stability AI, DeviantArt and Midjourney) for infringing copyright in the development of artistic works created by AI. Getty Images, one of the world’s largest agencies, has also taken legal action in London’s High Court of Justice against Stability AI, alleging that the company infringed intellectual property rights, including copyright, in the content it represents.
