IP and AI the ChatGPT case and the recent decision of the Italian Privacy Guarantor 2023-04-14
ChatGPT is a self-generating chat that collects a vast amount of data and processes it using Natural Language Processing, extracting it from textbooks, newspapers, websites and different articles to make it understandable and similar to human language.
ChatGPT, through programs, can summarise large amounts of data or respond quickly to voice commands.
The content created by ChatGPT is derived from materials previously generated by other authors, and uncertainties arise precisely from the copyright implications of reusing them.
A ChatGPT user may thus unwittingly end up infringing the author's copyright if they publish output that looks too much like an existing work.
At the same time, the most critical issues concern artists who see artificial intelligence tools operating on significant amounts of their works, replicating their style.
In Europe, text and data mining (TDM-the process used by some AIs to scrap online content) does not constitute copyright infringement if performed within the existing legal framework. Still, copyright holders can disable it and refuse TDM on their content (see Article 4 of Directive 2019/790)₁
ChatGPT can be helpful in terms of intellectual property insofar as it can provide information on patents, trademarks, copyrights, and trade secrets and can provide results in IP law or case law research. It should be remembered that sources may come from other jurisdictions, such as the United States, whose patent/trademark/copyright laws differ from the laws applicable in the European Union.
Artificial intelligence systems can be used both to create inventions protectable under patent law and to identify potential intellectual property infringement; thus, they are as much a tool for creation as they are for practical application, both of which, however, have confidentiality and privacy implications for the users themselves.
It should be noted that on the latter aspect, the Privacy Guarantor in Italy, by Order No. 112(₂)of March 30, 2023, has ordered the blocking of ChatGPT until it complies with privacy regulations, temporarily limiting the processing of Italian users' data and at the same time apprehending an investigation.
Sources:
1. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019L0790&from=EN
2. https://www.garanteprivacy.it/home/docweb/-/docweb-display/docweb/9870832
Back
ChatGPT, through programs, can summarise large amounts of data or respond quickly to voice commands.
The content created by ChatGPT is derived from materials previously generated by other authors, and uncertainties arise precisely from the copyright implications of reusing them.
A ChatGPT user may thus unwittingly end up infringing the author's copyright if they publish output that looks too much like an existing work.
At the same time, the most critical issues concern artists who see artificial intelligence tools operating on significant amounts of their works, replicating their style.
In Europe, text and data mining (TDM-the process used by some AIs to scrap online content) does not constitute copyright infringement if performed within the existing legal framework. Still, copyright holders can disable it and refuse TDM on their content (see Article 4 of Directive 2019/790)₁
ChatGPT can be helpful in terms of intellectual property insofar as it can provide information on patents, trademarks, copyrights, and trade secrets and can provide results in IP law or case law research. It should be remembered that sources may come from other jurisdictions, such as the United States, whose patent/trademark/copyright laws differ from the laws applicable in the European Union.
Artificial intelligence systems can be used both to create inventions protectable under patent law and to identify potential intellectual property infringement; thus, they are as much a tool for creation as they are for practical application, both of which, however, have confidentiality and privacy implications for the users themselves.
It should be noted that on the latter aspect, the Privacy Guarantor in Italy, by Order No. 112(₂)of March 30, 2023, has ordered the blocking of ChatGPT until it complies with privacy regulations, temporarily limiting the processing of Italian users' data and at the same time apprehending an investigation.
Sources:
1. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019L0790&from=EN
2. https://www.garanteprivacy.it/home/docweb/-/docweb-display/docweb/9870832