Leaked European Union documents point to heavy restrictions on artificial intelligence (AI) technology according to a draft of their upcoming rules.
The draft outlines the banning of “high-risk” AI systems and strict requirements for use approval. Companies that do not comply to the standards outlined could be subject to a fine of as much as 20 million euros ($23.9 million) or 4% of their turnover.
If implemented, these rules would be the most restrictive the AI industry has ever seen in what the European Union has called a “human-centric” approach. This would affect systems currently in place that perform tasks like handing out social security benefits, visa applications, or credit assessments.
These would be considered “high-risk” by the new proposed rules set.
The proposal will ban social scoring systems, like the ones used in China, and has determined that they “contravene the Union values”. The rules would also outright ban any AI systems that manipulate people’s behavior or provide any form of mass surveillance to the government.
The rules do provide one exception, stating that police authorities may be authorized to use mass surveillance AI “if they are fighting serious crime”.
© Fourth Estate
® — All Rights Reserved.
This material may not be published, broadcast, rewritten or redistributed.