6 March European parliamentarians praised the organization Regulation on artificial intelligence, however, critics inform that it is good for companies and contains shortcomings in crucial areas.
Daniel Leufer, elder Policy Analyst at the Brussels office of the Organization for Digital civilian Rights Access Now, called EU landmark act on AI "a failure from the human rights position and a triumph for the technology and police industry".
The first specified comprehensive regulation in the planet on rapidly evolving technology was adopted by members of the European Parliament after December negotiations on the finalisation of the act by 523 votes to 46 (49 abstentions). After performing the essential formal activities, the provisions are to enter into force in May or June and the individual provisions will apply within respective years.
"Although the adoption of the world's first rules on the improvement and usage of artificial intelligence is simply a milestone, it is disappointing that the European Union and the 27 associate States have decided to put manufacture and law enforcement interests on the protection of people and their rights" - summarized Mher Hakobyan, AI advisor at Amnesty International.
Derogations to manufacture lobbyists
The Act applies a risk-based approach to products and services utilizing artificial intelligence. Associated Press Agency informs:
‘In line with the assumptions, most systems utilizing AI, specified as algorithms to urge content or spam filters, are expected to be at low risk. Companies can decide whether they want to comply with the requirements and codes of conduct. The usage of high-risk AI, e.g. in medical equipment or critical infrastructure specified as a water or electricity network, will gotta meet stricter requirements, including the usage of advanced quality data and supply transparent information to users.
Some uses of AI have been prohibited due to unacceptable risks. These are social evaluation systems that regulate human behavior, any types of prevention systems and emotional designation systems in schools and workplaces. The ban besides includes police face scanners in public places utilizing distant biometric recognition systems utilizing AI, but for serious crimes specified as kidnapping or terrorism.’
While any praise common sense regulations and safeguards, Leufer believes that “the fresh AI Act is full of concessions to manufacture lobbyists and exceptions to the most dangerous applications of artificial intelligence by law enforcement and immigration authorities. In turn, bans abound in loopholes and do not truly block any of the most dangerous uses of this technology."
Hakobyan is afraid about how the fresh regulation will affect the situation of immigrants, refugees and asylum seekers, and emphasisesthat ‘it does not prohibit reckless the usage and export of draconian AI technologies.’
Strengthening Europe
Access Now and Amnesty global are part of the #ProtectNotSurveil coalition, which published a joint message informing that the AI Act "precises a dangerous precedent", in peculiar as regards the exclusion of law enforcement, immigration and national safety agencies.
The coalition besides includes organisations specified as EuroMed Rights, European Digital Rights and Statewatch. manager of the last one, Chris Jones, states in a message that “the act may and is new, but fits into an older communicative in which EU bodies and agencies, including Frontexhave been violating the rights of immigrants and refugees for decades."
Frontex, officially the European Border and Coast defender Agency, has long been subject to criticism from human rights groups, in view of their behaviour towards people entering the EU, especially by sea.
"With a number of fresh restrictive asylum and migration rules, the AI Regulation leads to the usage of digital technologies in order to strengthen the “Fortress of Europe” and to reduce the number of people coming to it in search of safety in fresh harmful ways", Jones warns. "Civil society groups across Europe and beyond should work together to offset the worst effects of these rules and proceed to make societies that supply surveillance and criminalisation care".
Much depends on national regulation
Laura Lazaro Cabrera, lawyer and manager of the program Equity and Data, conducted within the European branch of the Center for Democracy and Technology, writesthat the act "will become a mention point for AI government around the world, as we are faced with a race against a time in which decision-makers gotta face forward the improvement of technologies with far-reaching consequences for fundamental human rights".
Cabrera emphasisesthat, following its vote, "many depends on the implementation of the act. erstwhile the media sound around it has stopped, we will gotta face the hard task of applying these complex, comprehensive and unprecedented rules in practice. It will not be without close cooperation with experts and citizens to guarantee that the explanation and application of the rules meets the objectives expressed in the act itself: it defends human rights, democracy and the regulation of law."
Karolina Ivanska of the European Center for Not-for-Profit Law speaks in a akin tone“Let us be clear: the act has serious deficiencies in terms of fundamental human rights, and in this respect it should not be regarded as a model for the AI legislation. Nevertheless, we will effort to implement it in the best possible way."
Yannis Vardakastanis, president of the European Disability Forum, in a message writes"The AI Act addresses the issue of human rights, but not as much as we expected – we call on the European Union to fill this gap with future initiatives".
Hakobyan of Amnesty global points out that "non-EU countries should draw lessons from the failure of the block to regulate AI technology decently and not be subject to force from the technology manufacture and law enforcement authorities at the phase of drafting legislation. Individual countries should enact solid and binding government on AI, where people and their rights are most important."
**
Jessica Corbett – editor of "Common Dreams".
She postponed Anna Opara.
