In Europe, a bill is being promoted that would impose ‘sufficient disclosure obligations’ on big tech that uses copyrighted works for AI learning. In response to this, Big Tech is being criticized for avoiding responsibility while infringing on copyright itself. In response, the European Union (EU) has proposed amending its AI law, arguing that the bill should include a provision requiring disclosure of detailed information about copyright-protected data when used.
This proposal is intended to hold Big Tech accountable for what data it uses and how it uses it. Of course, it is questionable whether these modifications will actually be made and whether the effects after modification can be guaranteed, but the fact that such modification proposals have already been made is meaningful, explained the professor at this university. This legislative trend can be interpreted as progressing along with concerns about changing AI due to the emergence of generative AI such as ChatGPT.
In France, it has recently been required that copyrighted works must be used before using them for AI learning. An amendment to the intellectual property law has been proposed that would require permission from authors. This amendment includes content that allows copyright holder organizations to receive compensation even when an individual’s data is copied to another party, such as an AI developer. This legislative move could be interpreted as an attempt to limit the broader use of data by AI developers.
AI learning data includes not only news from media outlets but also posts posted by individuals on blogs or cafes, so this amendment means that it will be difficult to solve the problem simply through legal battles with companies.