Mario Draghi, the former president of the European Central Bank, has called for a pause on EU artificial intelligence regulation, just as he did with sustainability reporting rules.

Though the AI Act was passed last year, Draghi has said in a speech that the laws are a “source of uncertainty” and should be put on hold until its “drawbacks” can be assessed.

Under the act, AI applications are assessed in four categories from “unacceptable” to “minimal risk”. Corporate users, or “deployers” of AI tools had a responsibility to conduct risk assessments before rolling new technology and an obligation to ensure data was free from bias.

The AI Act has been a major source of discussion as companies look to accelerate its use in a bid to improve competitiveness and gain efficiencies.

However, when Draghi published his competitiveness report for the EU last year, he identified the AI Act, along with other legislation, as helping create regulatory uncertainty and a risk of “exclusion” from innovation for EU companies.

‘Understand the drawbacks’

In his speech, according to euronews.com, Draghi says it is the next stage of AI Act implementation that could cause issues. “In my view, implementation of this should be paused until we better understand the drawbacks,” he said.

It was the same report last year that caused the EU to rethink the introductions of corporate sustainability reporting. In that report, Draghi wrote: “The EU’s sustainability reporting and due diligence framework is a major source of regulatory burden magnified by a lack of guidance to facilitate the applications of complex rules and to clarify the interaction between various pieces of legislation.”

In an article for Board Agenda, cybersecurity and AI expert Kamal Bechkoum wrote that the AI Act is “redefining compliance obligations”. He added: “Companies that are unprepared for these changes face legal and reputational damage.”

There has already been talk of delaying the AI Act. There have been demands that the Act should wait for the release of technical standards or include more exemptions and perhaps even introduce “waivers” for “low-complexity” AI from the assessment regime.

Law firm DLA Piper wrote back in June: “For most businesses, the list of prohibited AI practices is unlikely to materially impact day-to-day activity.

“However, AI use cases categorised as high-risk cover a broader array of typical business activities, including AI used in recruitment or job allocation and AI used in credit checking or scoring.

“Therefore, those racing to implement appropriate controls for such systems will welcome any prospect of additional time.”