Big Tech monopolises AI

Big Tech’s last-minute blitz is further diluting the AI Act - New publication shows how

Scandals from the Netherlands to Australia have highlighted the severe risks to fundamental rights posed by AI systems. The EU Artificial Intelligence Act is an attempt to rein in some of these risks.

"Byte by byte: How Big Tech undermined the AI Act”, new research by Corporate Europe Observatory, shows, however, how Big Tech has been trying to further dress down the AI Act during the last stretch of negotiations.

To gain access to every institution involved in the trilogue process, Big Tech executives have undertaken a lobbying blitz, meeting EU commissioners and EU heads of state ahead of trilogue meetings. As a result, negotiations on the AI Act are at risk of being derailed.

In its new research, Corporate Europe Observatory collected new numbers showing that in 2023 only, 66% of meetings held with MEPs on the AI Act were with corporate interest. Meetings with European Commission's high-level officials were even more skewed towards industry lobbying - 86% were with corporate interest. In one document obtained through Freedom of Information, Meta sent 134 pages of detailed changes and amendments to EU governments on the AI Act.

Big Tech has focused its lobbying efforts on the most advanced AI models, the so-called "foundation models", including ChatGPT. While the European Parliament has introduced some requirements to regulate foundation models specifically, Big Tech has pushed lawmakers to leave these systems unregulated.

Big Tech's lobbying push is proving to be successful. Already, EU institutions agreed that tech companies can self-assess if an AI system poses a high risk. While tense negotiations are still ongoing on the regulation of foundation models, major member states, under pressure from Big Tech and their own industry, are pushing to leave the most advanced AI models unregulated.

Bram Vranken, Corporate Europe Observatory Researcher and Campaigner, says:
“Big Tech's last-minute lobbying offensive threatens to derail the AI Act and to jeopardise our fundamental rights when faced with opaque and unaccountable AI systems.

Undue industry interference is not just threatening the AI Act, it also undermines democratic decision-making. It is time to restrict Big Tech's access to the EU institutions.”

ENDS

For more information, contact: Bram Vranken, Corporate Europe Observatory Researcher and campaigner:
bram@corporateeurope.org

Notes to editor:

This article continues after the banner

Support CEO so we can stay independent!