To make artificial intelligence safer, 18 countries have adopted an international agreement, including Germany. This is about more controls on providers. However, the agreement is non-binding.
The USA, Germany and 16 other countries have presented what appears to be the first detailed international agreement to protect against the misuse of artificial intelligence. In the 20-page document, the 18 countries agreed that companies designing and using AI must develop and deploy it in a way that protects customers and the general public from misuse.
“This is the first time we’ve seen confirmation that it shouldn’t just be about cool features and how quickly we can bring them to market,” said US Cybersecurity and Infrastructure Security Agency Director Jen Easterly told Reuters. Rather, the agreement’s guidelines would reflect that the most important thing to do in the design phase was security.
The agreement is non-binding
However, the agreement is non-binding and mainly contains general recommendations, such as monitoring AI systems for misuse, protecting data from manipulation and checking software providers. Still, Easterly says it’s important that so many countries embrace the idea that AI systems must put safety first.
The agreement is the latest initiative in a series of attempts by governments around the world to have more influence over the international development of AI. Europe is ahead of the United States in regulations surrounding AI. For example, legislators in France, Germany and Italy recently reached an agreement on how to regulate artificial intelligence, supporting mandatory self-regulation through codes of conduct for so-called basic models of AI.
In addition to the USA and Germany, the 18 countries that have signed the new guidelines include Great Britain, Italy, Czech Republic, Estonia, Poland, Australia, Chile, Israel, Nigeria and Singapore.