Europe and the United States propose a code of conduct for artificial intelligence
.png)
Although concerns about the impact of Artificial Intelligence (AI) models have already begun to move regulatory initiatives around the world, time marches on and offers and products increase with no legal framework in sight to address the potential problems of this technological advance.
As an alternative to rein in developers and mitigate risks while a regulation is being approved, on May 31 the European Union and the United States announced their intention to work together on the design of a code of conduct for AI companies, a draft of which would be presented in the coming weeks.
The announcement was made during a meeting in Sweden of the Trade and Technology Council formed by the European Union and the United States, in which Secretary of State Anthony Blinken was present. According to statements by Margrethe Vestager, European Commission vice-president for digital affairs, the project seeks to establish a series of best practices while a regulation is being developed. Governments in other parts of the world are expected to join the initiative.
"It's a way for democracies to respond in real time to issues that are really on our plate right now. I'm looking forward to working deeply and quickly with as many as possible," Vestager said.
Although the codes of conduct are voluntary in nature, some moves by Sam Altman, CEO of OpenAI and one of the most visible faces in the industry, would indicate that the industry - or at least its main players - would embrace these preliminary rules, either out of conviction or because of public pressure.
In mid-May, Altman appeared before the U.S. Congress, where, to the surprise of many, he asked lawmakers to regulate his own business. For Natasha Lomas, a journalist at TechCrunch, it would be very hypocritical if, after having acted in this way, his company did not commit to voluntary agreements.
As this is just an initial idea, there is still no information on what kind of behaviors AI companies should comply with under this code. In any case, Alexandra Reeve Givens, president of the Center for Democracy and Technology, who was also present at the meeting in Sweden, called attention to some harmful events that are already occurring from these technologies, such as deep fakes that could affect the course of an election and impersonations and scams through voice tools.
Givens also believes it is important to establish transparency standards and set a common framework that can serve regulators in other parts of the world, civil society and marginalized communities that are more exposed to the risks of this technology.
Codes of conduct - which have been implemented previously for social networking companies - constitute a model of co-regulation in which, although there are no strict obligations, companies commit to comply with certain agreements, while retaining their freedom to carry out their activities according to their own criteria.
For Fabiana Di Porto and Marialuisa Zuppeta, researchers at the University of Salento, this type of model may lack real force because, since it is a voluntary matter, the only punishment for non-compliance is reputational.
This lack of strength can be especially noticeable in scenarios where companies do not consider their reputation as one of their main assets. This is the case of Twitter in the hands of Elon Musk, which recently withdrew from the European Union's Code of Best Practices on Disinformation, which it had joined in 2018.
In any case, this case also shows how codes of conduct can be a backstop while a law comes into force, since, as explained by European Commissioner Thierry Breton, Twitter will only be freed from its duties for a few months, given that from August these voluntary provisions will become mandatory in light of the Digital Services Act.
