Sources: EU regulators are struggling to gather enough evidence to bring antitrust charges against Amazon despite working on the case for nearly two years
Regulators face difficulties gathering evidence against tech giant nearly two years after probe was announced
Context & Ripple Effects
Two years into the EU's probe over how Amazon uses marketplace data, Brussels cannot assemble enough admissible evidence to bring formal charges — a stalemate that shapes everything downstream in this coverage thread. Rather than litigate a conduct case it may lose on proof, the Commission ends up extracting concessions instead: Amazon agrees in mid-2022 to share more data with rivals and widen product choice, which becomes the template for the final deal closing both investigations that December.
The enforcement gap does not go unfilled: within months of this report, the UK's competition watchdog announces a formal investigation explicitly mirroring the EU probe, keeping pressure alive on exactly the data-use theory Brussels struggled to prove.
First-order effects
- Amazon faces no formal Statement of Objections despite the two-year probe, shifting the battleground from courtroom charges to negotiated behavioral commitments over rival-product visibility and non-Prime offers.
- The European Commission must choose between closing the case and accepting voluntary remedies, ceding the enforcement initiative to other authorities — most immediately the UK watchdog, whose planned formal investigation replicates the EU's data-use theory.
Second-order effects
- Negotiated commitments become the de facto outcome: Amazon concedes data-sharing and wider product choice to end the probes, giving competitors better marketplace visibility without any finding of wrongdoing.
- With the EU case resolved by settlement rather than precedent, national regulators step in — the UK's mirror probe means Amazon faces duplicate scrutiny of the same conduct, raising compliance costs across jurisdictions.
Third-order effects
- The Commission's playbook shifts toward venues where evidence is easier to marshal: merger reviews like the four-month Phase 2 investigation into Amazon's $1.7B iRobot acquisition, rather than hard-to-prove conduct cases.
- The same evidentiary threshold recurs at the frontier of AI deals — the EC drops its merger review of Microsoft's OpenAI investment for lack of evidence of control, then explores an antitrust angle on the same deal, suggesting structural review will keep absorbing cases conduct rules can't carry.
The trend: EU competition enforcement against Big Tech is migrating from conduct charges that stall on evidence toward negotiated behavioral commitments and aggressive merger scrutiny.