Pressure Mounts: Lobbyists Urge EU Authorities to Delay Implementation of AI Act Amid 3.4 Trillion Euro Stakes

The technology association CCIA Europe has formally requested EU leaders to postpone the implementation of the new provisions of the AI Act, originally set to take effect on August 2. This legislation is a comprehensive European law aimed at regulating artificial intelligence. Lobbyists from the group, which includes major companies like Alphabet, Apple, and Meta* (designated as an extremist organization in Russia), have expressed concerns that the European Commission has yet to release the final clarifications on the legislation. As a result, industry representatives are unclear on how to proceed. A survey conducted by AWS and Strand Partners indicates that 68% of industry professionals lack understanding of the obligations imposed by the AI Act.

The stance of CCIA Europe is echoed by some European politicians. For example, Swedish Prime Minister Ulf Kristersson has described the regulations as «confusing» and suggested a pause in their implementation. If the European Commission does not grant a delay, companies will face uncertainty as they navigate compliance, risking penalties of up to €35 million or 7% of global revenue as outlined in the AI Act. This uncertainty could jeopardize the European Union’s ambition to leverage artificial intelligence for an increase in GDP by €3.4 trillion by 2030.

Currently, the European Commission remains steadfast in its position. Digital Policy Commissioner Henna Virkkunen assured lawmakers, «We aim to implement the law in an innovation-friendly manner, but a pause is not on the table.» The Council of the EU may discuss this issue at the upcoming summit in Brussels in July.

The AI Act is set to take effect on August 1, 2024, and includes a phased introduction of regulations for the AI industry. A complete ban on «unacceptable» AI practices, such as social scoring and deceptive manipulative interfaces, will come into effect on February 2, 2025. Starting August 2, 2025, the law will begin regulating general-purpose AI models—this is the phase CCIA Europe is requesting to delay. By this date, documentation, model cards, and training datasets are expected to be standardized; plans are also in place to establish a European AI office and national authorities. Additionally, all high-risk systems, such as those used in finance, HR, and healthcare, must undergo assessment by August 2, 2026. A final deadline of August 2, 2027, has been set for existing models in Europe to either comply with AI Act requirements or exit the market.

P.S. You can support me by subscribing to the channel «the runaway neural network«, where I share insights about AI from a creative perspective.