Few words have echoed more persistently through Brussels than “simplification.” Since presenting her key political priorities in July 2024, European Commission President Ursula von der Leyen has placed regulatory simplification at the heart of her mandate, promising measures that would “radically lighten the regulatory load and related costs for people, businesses, stakeholders, and public administrations in the European Union.”
The ambition was never about deregulating or suppressing the EU’s digital rulebook. The simplification agenda was a targeted exercise intended to identify and eliminate legislative overlaps, reduce disproportionate administrative burdens, and create a regulatory environment that works for companies of all sizes, particularly small and medium-sized enterprises (SMEs). As recently as last week, the 27 heads of state and government reaffirmed that commitment, calling for continued efforts “to ambitiously simplify rules and reduce administrative burdens … to ensure a regulatory framework that is both innovation- and SME-friendly.”
Yet when ambition had to translate into action — specifically in the context of the Artificial Intelligence (AI) Omnibus proposal — the outcome fell far short of the promise. What was billed as “unprecedented” simplification now risks becoming little more than a symbolic gesture.
What Went Wrong?
The Commission’s initial proposal was, in fact, a model of restraint: short, targeted, and disciplined, introducing only a limited series of amendments to the EU AI Act.
The same cannot be said of the European Parliament and the Council, whose actions either significantly diluted or reversed several of the Commission’s key elements. This includes the reintroduction of the obligation to register high-risk AI systems used solely for preparatory tasks (Article 6), and the near-total restriction of provisions allowing personal data processing for bias detection and mitigation to only the narrowest of circumstances (new Art. 4a).
The European Parliament went further still, reducing the proposed grace period for transparency requirements applicable to deepfakes (Art. 50) from six months to three, and, in direct contradiction with the stated objective of simplification, added new documentation requirements (Art. 25) rather than removing existing ones.
Does this pattern sound familiar? Those with long institutional memories will recall the “Better Regulation” agenda under the Juncker Commission, an initiative that generated considerable political energy before quietly dissolving into procedural complexity. Simplification cannot be allowed to follow the same trajectory, chasing a compelling headline that yields an empty shell in practice.
What Must Happen Now
As trilogue negotiations commence, the three EU institutions bear the responsibility of what comes next: Either realign with the original ambition embodied in the Commission’s proposal, or accept that this exercise will fall short of its stated purpose.
Only a genuinely ambitious and unprecedented simplification effort will ease the regulatory burden on businesses, while promoting innovation and reinforcing Europe’s competitiveness. If Europe fails to make good on simplification, and businesses operating across the EU face a more challenging path to compliance, the AI Omnibus risks becoming precisely what it was designed to prevent: A missed opportunity dressed up as reform.
