The President of the European Commission, Ursula von der Leyen, called on Wednesday for the creation of a “global panel” of experts to assess the risks of artificial intelligence for humanity, based on the model of the Group intergovernmental experts on climate change (IPCC). “Hundreds of leading developers, academics and experts in the field of AI have recently alerted us” to the risk of humanity’s extinction, recalled the president of the European executive, in a speech to the European Parliament in Strasbourg.
Reducing this risk “should be a global priority, along with other risks threatening society as a whole, such as pandemics and nuclear war,” she said. Artificial intelligence “is evolving more quickly than its designers themselves anticipated. We therefore have little time left to guide this technology in a responsible manner,” she worried, calling for the establishment of a “new global framework for AI, based on three pillars: safeguards, governance and the direction of innovation. “We should work together with our partners to all have the same understanding of the impact of AI in our societies,” she said, citing “the invaluable contribution of the IPCC for the climate, this global panel which provides policy makers with the most recent scientific data.
“I think we need a similar body for AI, in terms of its risks and benefits to humanity. This body would bring together scientists, technology companies and independent experts. This will enable us to develop rapid and coordinated action on a global level,” said Ursula von der Leyen.
The distribution on social networks of false images, larger than life, created from applications like Midjourney, has alerted to the dangers of artificial intelligence systems for democracy. The potential of these technologies, which can also write dissertations, compose music or facilitate medical diagnoses, fascinates as much as it worries. The EU hopes to conclude the world's first regulation to regulate these innovations before the end of the year. However, it will not come into force before 2026. Future legislation should notably prohibit mass surveillance applications for citizens, and impose standards on a list of services considered at risk.
It is currently the subject of negotiations between the 27 member countries of the EU and the European Parliament. “Our AI legislation is already a model for the whole world. We must now adopt the rules as soon as possible and move on to their implementation,” asked Ursula von der Leyen.