Generative AI developers will be required to disclose any copyrighted material used to train algorithms. The developers of the Law on Artificial Intelligence in the European Union came to such an agreement, writes Reuters.
According to a source familiar with the situation, the regulation was a late addition drawn up over the past two weeks. Some members of the committee initially proposed to completely ban the use of licensed materials for training generative algorithms. However, this was abandoned in favor of a requirement for transparency.
“Despite conservative desires for increased surveillance and left-wing fantasies of over-regulation, Parliament has found a solid compromise. It will allow proportional regulation of AI, protect the rights of citizens, as well as promote innovation and stimulate the economy,” said MEP Svenja Khan.
Macquarie analyst Fred Havemeyer said the EU proposal was “tactful.” At the same time, some suggested the approach of “prohibit first, ask questions later”.
“The EU is at the forefront of AI regulation,” he told Reuters.
The European Commission has started drafting a bill to regulate AI in 2021. Since then, generative AI algorithms like ChatGPT have become especially popular.
Members of the European Parliament agreed to bring the document to the next stage, during which they will process the final details of the bill.
According to the proposals, AI tools will be classified according to the perceived level of risk, from minimal to unacceptable. Areas of concern may include biometric surveillance, the spread of misinformation or discriminatory language.
High risk instruments will not be banned. However, their users will be required to maintain maximum transparency.
The latest amendment will oblige developers of generative systems like ChatGPT or midjourney disclose any copyrighted material used to train their algorithms.
Earlier media reported that the European Parliament was close to agreeing on proposals for the Law on Artificial Intelligence.
Recall that in April, a group of European legislators announced plans to develop draft rules to control the development of AI systems.
Found a mistake in the text? Select it and press CTRL+ENTER
Cryplogger Newsletters: Keep your finger on the pulse of the bitcoin industry!
Generative AI developers will be required to disclose any copyrighted material used to train algorithms. The developers of the Law on Artificial Intelligence in the European Union came to such an agreement, writes Reuters.
According to a source familiar with the situation, the regulation was a late addition drawn up over the past two weeks. Some members of the committee initially proposed to completely ban the use of licensed materials for training generative algorithms. However, this was abandoned in favor of a requirement for transparency.
“Despite conservative desires for increased surveillance and left-wing fantasies of over-regulation, Parliament has found a solid compromise. It will allow proportional regulation of AI, protect the rights of citizens, as well as promote innovation and stimulate the economy,” said MEP Svenja Khan.
Macquarie analyst Fred Havemeyer said the EU proposal was “tactful.” At the same time, some suggested the approach of “prohibit first, ask questions later”.
“The EU is at the forefront of AI regulation,” he told Reuters.
The European Commission has started drafting a bill to regulate AI in 2021. Since then, generative AI algorithms like ChatGPT have become especially popular.
Members of the European Parliament agreed to bring the document to the next stage, during which they will process the final details of the bill.
According to the proposals, AI tools will be classified according to the perceived level of risk, from minimal to unacceptable. Areas of concern may include biometric surveillance, the spread of misinformation or discriminatory language.
High risk instruments will not be banned. However, their users will be required to maintain maximum transparency.
The latest amendment will oblige developers of generative systems like ChatGPT or midjourney disclose any copyrighted material used to train their algorithms.
Earlier media reported that the European Parliament was close to agreeing on proposals for the Law on Artificial Intelligence.
Recall that in April, a group of European legislators announced plans to develop draft rules to control the development of AI systems.
Found a mistake in the text? Select it and press CTRL+ENTER
Cryplogger Newsletters: Keep your finger on the pulse of the bitcoin industry!