The United States and the European Union share a fascination, and concern, for all the possibilities—and risks—involved in the rapid advance of generative artificial intelligence (AI) across the board. Europe is already preparing with a pioneering law that seeks to regulate all aspects of AI, but until it is approved – it is still in the legislative negotiation phase in Brussels – and enters into force, it may still be years. Precious years in which this type of technology can continue to advance without control. This is what Washington and Brussels want to avoid, which intend to soon present a voluntary "code of conduct" for companies that develop this type of technology, as advanced on Wednesday at a bilateral meeting in Sweden.
"In the coming weeks, we will advance a proposal for a code of conduct for generative AI to which the industry can commit on a voluntary basis," said the European Commission's vice president for digital and competition, Margrethe Vestager, at the end of a ministerial meeting of the sector with the United States in Lulea, Sweden. The text, prepared together with the United States and for which both the industry and experts will be consulted, should be "very, very soon, in the coming weeks," he added.
The announcement comes 24 hours before Sam Altman, the founder of the company OpenAI creator of ChatGPT, the technology that has set off all the alarms, arrives in Brussels as part of a European tour that has taken him to several capitals, including Madrid. At EU headquarters, Altman, who is one of the proponents of a regulation of generative AI, will meet with the president of the European Commission, Ursula von der Leyen, in a "working meeting" closed to the press. But first, Altman met with Vestager and U.S. Commerce Secretary Gina Raimondo in Lulea, where they discussed the possible voluntary code of conduct, which according to "ideas" discussed with generative AI developers, could include controls such as watermarking, or external audits, Vestager said in a tweet.
⚠️Accountability on #AI can't wait. It is NOW. Today #TTC kicked off work on a 1st voluntary AI #CodeOfConduct. We'll work with our key partners & the #AI community on #safeguards to make AI responsible, safe & trustworthy. This is a huge step in a race we can't afford to lose. pic.twitter.com/WBcazIysiK
— Margrethe Vestager (@vestager) May 31, 2023
The EU wants to be a pioneer in the regulation of Artificial Intelligence, for which it has made a legislative proposal, the AI Act. The text should be approved at the next plenary session of the European Parliament, in mid-June, after which the negotiation stage will begin with the Council of the EU and the Commission to reach an agreement on a final text that, after its new ratification by the Twenty-seven and the European Parliament, will enter into force throughout the EU.
But as Vestager noted Wednesday, generative AI is a game-changer "so powerful" that it may not be possible to wait for lawmakers to finish their work. "We have several different legislative procedures, they will take 2 or 3 years before they come into force, and we are talking about an incredible technological acceleration," Vestager explained in an appearance with US Secretary of State Anthony Blinken and Raimondo.
A voluntary code of conduct, which Vestager says will be joined by companies from other countries, such as Canada, the United Kingdom, Japan or India, could be a regulatory "bridge" until the current regulations come into force – or future laws in other regions – and thus give citizens confidence that "democracies respond" to their challenges and concerns.
Watermarking, external audits, feedback loops - just some of the ideas discussed with @AnthropicAI and @sama @OpenAI for the #AI #CodeOfConduct launched today at the #TTC in #Luleå @SecRaimondo Looking forward to discussing with international partners. pic.twitter.com/wV08KDNs3h
— Margrethe Vestager (@vestager) May 31, 2023
This move comes just one day after a group of 350 executives, researchers and engineers expert in this technology signed an open letter of only 22 words about the risk of this technology. "Mitigating the extinction risk [for humanity] of AI should be a global priority along with other risks on a societal scale, such as pandemics and nuclear war," cites the statement that has been signed, among others, by the senior executives of three major artificial intelligence companies: Sam Altman (executive chairman of OpenAI), Demis Hassabis (Google DeepMind) and Dario Amodei (Anthropic). Signatories also include researchers Geoffrey Hinton and Yoshua Bengio, who are often considered godfathers of the modern AI movement. Hinton left Google a few weeks ago, where he occupied a vice presidency, because he believes that this technology can lead us to the end of civilization in a matter of years, as he confessed to EL PAÍS.
The statement comes at a time of growing concern about a rapidly developing sector that is difficult to control. Sam Altman himself had already spoken out on this issue during his appearance before the US Senate, when he recognized the importance of regulating generative artificial intelligence. "My worst fear is that this technology will go wrong. And if it goes wrong, it can go very wrong," he said just two weeks ago during the first AI hearing on Capitol Hill. The father of OpenAI added that he understood that "people are eager about how [AI] can change the way we live," and that for this reason it is necessary to "work together to identify and manage the potential disadvantages so that we can all enjoy the tremendous advantages."
Nor was it the first time that one of the entrepreneurs most involved in this technology made statements of this magnitude regarding the future of AI. In March, more than a thousand intellectuals, researchers and entrepreneurs had signed another open letter in which they asked to stop for "at least six months the development of AI systems more powerful than GPT4", the latest version of ChatGPT. In the letter, the signatories warned that OpenAI's tool is already capable of competing with humans in a growing number of tasks, and that it could be used to destroy jobs and spread disinformation.
You can follow EL PAÍS Tecnología on Facebook and Twitter or sign up here to receive our weekly newsletter.