European Union politicians have agreed to draft artificial intelligence (AI) rules that include a blanket ban on the use of the technology in biometric surveillance and predictive policing systems.
Key points:
- The EU Commission's proposed AI Act will regulate the use of artificial intelligence
- The rules include a ban on AI for real-time biometric surveillance, emotion recognition, and also predictive policing
- EU countries are expected to push back against limits on biometric identification systems
The EU Commission's proposed landmark laws, which were amended after a vote on Wednesday, are designed to set a global standard for protecting people from the technology, following warnings from experts that AI presents an "extinction" risk similar to that of pandemics and nuclear war.
The rapid adoption of Microsoft-backed OpenAI's ChatGPT and other bots has led top AI scientists and company executives — including Tesla's Elon Musk and OpenAI's Sam Altman — to call for coordinated international regulation of generative AI.
Member of the European Parliament, Brando Benifei, has been in charge of negotiating the draft of the future AI Act, which has been in the works since 2021 and is expected to comprise the world's first comprehensive AI laws.
"While big tech companies are sounding the alarm over their own creations, Europe has gone ahead and proposed a concrete response to the risks AI is starting to pose," he said.
Mr Benifei said he was confident the draft rules balanced fundamental rights and technological innovation with providing legal certainty to businesses.
"We are on the verge of putting in place landmark legislation that must resist the challenge of time," he said.
"It is crucial to build citizens' trust in the development of AI, to set the European way for dealing with the extraordinary changes that are already happening, as well as to steer the political debate on AI at the global level."
Among other rules, the EU wants to ban any "intrusive and discriminatory uses of AI systems", such as the use of real-time and remote biometric identification systems, emotion recognition, and predictive policing systems based on profiling, location or past criminal behaviour.
The EU also wants to ban the indiscriminate scraping of facial images from social media or CCTV footage to create facial-recognition databases.
Generative tools such as ChatGPT would have to comply with additional transparency requirements, such as disclosing that content is AI-generated and publishing summaries of copyrighted data used to train the model.
Technological companies Microsoft and IBM welcomed the draft AI rules but looked forward to further refinement of the proposed legislation.
"We believe that AI requires legislative guardrails, alignment efforts at an international level, and meaningful voluntary actions by companies that develop and deploy AI," a Microsoft spokesperson said.
EU politicians will now have to negotiate the details with the EU's 27 countries before the draft rules can become legislation.
The battle over biometrics
The AI rules around the use of biometrics — data relating to human characteristics — are the most controversial limitations under debate, with EU countries expected to push back against them.
The centre-right European People's Party has argued the technology could be of vital importance in combating crime and in counter-terrorism intelligence, as well as in searches for missing children.
While most big tech companies acknowledge the risks posed by AI, others such as Meta, which owns Facebook and Instagram, have dismissed warnings about the potential dangers.
"AI is intrinsically good because the effect of AI is to make people smarter," Meta's chief AI scientist Yann LeCun said at a conference in Paris on Wednesday.
Under the proposed laws, AI systems that could be used to influence voters and the outcome of elections — as well as recommender systems used by social media platforms with over 45 million users — would be added to the high-risk list.
It would mean social media giants Meta and Twitter would fall under that classification.
"AI raises a lot of questions — socially, ethically, economically. But now is not the time to hit any 'pause button'. On the contrary, it is about acting fast and taking responsibility," EU industry chief Thierry Breton said.
Mr Breton said he would travel to the United States next week to meet Meta CEO Mark Zuckerberg and OpenAI's Mr Altman to discuss the draft AI Act.
ABC/Reuters