Vulcan Insight | European Parliament adopts its position on the AI Act

On Wednesday earlier this week, Members of the European Parliament (MEPs) gathered in Strasbourg for this month’s plenary session, with the AI Act being very high on the agenda. From its initial proposal in April 2021 and following many months of negotiations between the two responsible parliamentary committees – the Committee on the Internal Market and Consumer Protection (IMCO) and the Committee on Civil Liberties, Justice and Home Affairs (LIBE) – the European Parliament’s plenary finally passed its version of the Artificial Intelligence Act (AI Act) after two years of wrangling.

Following this week’s European Parliament (EP) approval of the AI act, amendments were presented to limit the risk involved with the use of AI. Changes proposed included requirements for manufacturers to rank their own AI technology between low and high risk before introducing it to the market. However, no rules of AI use in the workplace were included in the act. The European Trade Union Confederation (ETUC) welcomed the improvements made by the EP, including consultation with workers and their unions before implementing and introducing AI to the workplace, carrying out assessments of the impacts that AI will have on fundamental human rights, and a clause to limit the use of AI systems to protect workers’ rights. A study conducted by the Organization for Economic Co-operation and Development (OECD) had shown that workers will benefit from positive impacts of AI if consultations with their trade unions have been conducted. However, the ETUC’s claims the EP failed to close a loophole that puts workers’ safety and rights at risk. Article 6 of the report states the AI will be restricted from the work environment if posing “significant risk”.

One point of contention was the Parliament’s proposal to ban facial recognition technology. The Parliament aims to curb the use of facial recognition in public spaces due to concerns about biases and its potential for mass surveillance. The European Commission’s initial proposal allowed exceptions for certain purposes, such as searching for missing children or dangerous criminals. However, the Parliament’s version removed these exceptions and expanded the ban’s scope, which raised concerns among some lawmakers who worried about hindering law enforcement. As Brando Benifei, the Co-Rapporteur from IMCO, announced at a press conference following the vote: ‘We have won in Parliament to maintain a clear safeguard — to avoid any risk of mass surveillance […] We need to negotiate with the governments, but today’s result gives us a stronger position.

The European Peoples Party (EPP) outlined the list of standards that need to be met with regards to a human-centred approach to AI based on European ethical and democratic values. Further claiming that AI is not to be feared but understood through regulating its risks, implementing “guardrail” would ensure a responsible development and use of powerful AI tools, such as chat GPT. Lead MEP Axel Voss equally supports the AI act in bringing benefits for citizens and “strengthening the European democratic values in the Global market.” Similarly, Shadow Rapporteur MEP Deirdre Clune believes the AI act is a “world first,” and further emphasised the importance in ensuring that such technology is developed and used responsibly whilst promoting and supporting innovation and economic growth.

The final version of the AI Act will now be negotiated in trilogues among lawmakers, Commission officials, and attachés from EU member countries. Some attachés have already stated that accepting a complete ban on facial recognition is not an option. The Parliament is determined to maintain safeguards against mass surveillance and intends to negotiate with governments to reach a consensus.

Compliments of Vulcan Consulting – a member of the EACCNY.