France to deploy AI surveillance ahead of the Paris 2024 Olympics

French authorities have accelerated the development of a dedicated law for the 2024 Paris Olympics, which would allow the use of AI video surveillance in public places. The Guardian reports.

The government says the proposed measures are necessary to ensure the safe conduct of events expected to attract 13 million spectators.

The law will enable authorities to automate video surveillance in which AI can detect suspicious activity in crowds. Algorithms will analyse images from stationary cameras or drones, highlighting ‘abnormal behaviour’. After this, the system will autonomously alert the police to potential violations.

AI systems will be used to monitor crowds both inside stadiums and in their vicinity. In addition, the algorithms could be deployed to monitor conditions on streets and in public transport.

French Sports Minister Amélie Oudéa-Castéra described the bill as introducing ‘necessary adjustments’ to ensure the Games go smoothly. Interior Minister Gérald Darmanin, for his part, said the new rules would provide a ‘basis’ for the Games’ security.

The proposal comes against the backdrop of the UEFA Champions League final in 2022, when mass disturbances occurred at the Stade de France; among the affected were children. Some attendees also reported that they had been robbed at the stadium.

The French government stressed that the safety proposals for the Olympics do not include the use of facial-recognition technologies. Ministers said that automated video surveillance had already been tested for some time.

Human rights groups have warned that France is seeking to use the Games as a pretext to widen police surveillance powers. They say the government proposal would cover all sporting, festive and cultural events from spring 2023 to June 2025. They fear the measure could become a permanent law.

Amnesty International described the proposal as a ‘watershed moment’ in the use of AI surveillance technologies in France. It is a ‘dangerous step’ for human rights, the rights defenders added.

‘We are deeply concerned […]. There is a problem defining suspicious activity: who will decide what behaviour is normal and what is not? […] We believe this proposal poses a danger to the right to privacy, freedom of expression, assembly and non-discrimination principles,’ said Katya Ru, a human rights and technology specialist at Amnesty International in France.

She added that the effectiveness of automated monitoring methods has not been demonstrated.

According to Amnesty International, the Games should not be used as a pretext for making extraordinary measures permanent.

Ru noted that previously states have used international sporting events as a pretext to pass laws that were later used for repressive purposes. As an example, the rights defender cited Russia’s introduction of facial-recognition surveillance during the 2018 FIFA World Cup.

A Bastien Le Querrec, spokesperson for the NGO La Quadrature du Net, also spoke out against the initiative. He said the government proposal ‘seems dangerous and disproportionate’.

Le Querrec said that until now in France a human must analyse CCTV footage. This makes constant surveillance of all public spaces impossible.

‘But with algorithms, this human limit no longer exists […]. Any person captured on camera could be subjected to behavioural analysis, and the algorithm would decide its normality,’ said the rights defender.

He added that it is unacceptable to use the Olympic Games for such political ends.

The French Senate has already approved the bill. In February, the lower house will consider it.

Earlier in October 2022, French authorities fined Clearview AI for GDPR violations.

Subscribe to ForkLog News on Telegram: ForkLog AI — all the news from the world of AI!