Five Years On: Big Hopes, Modest Results
When the Bundeskartellamt was handed the world’s first bespoke competition toolkit against the largest digital platforms in January 2021, it was billed as a genuine breakthrough. Section 19a of the Act against Restraints of Competition (GWB) allows the authority to designate a company as being “of paramount significance for competition across markets” and then, without first having to prove market dominance or actual harm in the traditional sense, prohibit specific conduct – such as self-preferencing or data leveraging – before the damage is done. To speed up things, the legislator even replaced the normal two-tier system of judicial review with a single-instance appeal.
Five years, a single prohibition decision and a fresh legislative reform later, it is worth asking how far that promise has been realised.
The big hopes
The rhetoric around the new provision was ambitious from the outset. Announcing the reform, Bundeskartellamt President Andreas Mundt framed it as a shift from reactive to preventive enforcement: the authority would now be able “to prohibit big tech companies from engaging in certain types of conduct much earlier and, so to speak, shut the stable door before the horse has bolted.” Mundt also called the new regime a “major step forward” that made Germany an “international pioneer,” at a time when the EU was still debating what would became the Digital Markets Act.[1]
Academic commentary at the time captured a mixture of ambition and unease. Jens-Uwe Franck and Martin Peitz published an article titled “Section 19a of the Reformed German Competition Act: A Too Powerful Weapon to Tame Big Tech”.[2] The concern, in other words was, that the new provision might prove too aggressive, not too weak.
A modest output
Measured against that billing, the enforcement record looks thin.
As of September 2026, the Bundeskartellamt has designated the world’s five largest tech companies – Alphabet, Amazon, Apple, Meta and Microsoft – as being of paramount significance for competition across markets, each for a period of 5 years. No surprise there. Importantly, though, unlike under the DMA, designation alone does not automatically (ipso iure) subject a company to additional restrictions. Obligations under Section 19a (2) bind a designated undertaking only once the Bundeskartellamt has individually ‘activated’ them, conduct by conduct, and obligation by obligation, through a formal decision prohibiting specific behaviour. Absent such a decision, a designated undertaking remains free to continue any conduct. Nor does Section 19a provide for private enforcement.
To date, the Bundeskartellamt has considered Section 19a (2) in seven investigations: three against Google, two against Amazon, and one against Meta and Apple. Of those seven, three[3] were opened under Section 19a; the other four had already been initiated under general abuse-of-dominance provisions of Article 102 TFEU/Section 19 GWB and were subsequently extended Section 19a. Of the three new investigations, two relied on Section 19a alone, while the third also relied on Article 102 AEUV.
By comparison, the EU Commission has opened twelve DMA investigations since March 2024 – in roughly half the time, despite the fact that the DMA does not require any additional administrative act to render the relevant obligations binding.
The outcomes of the seven Section 19a proceedings are even more modest. Only one has resulted in a formal prohibition decision under Section 19a(2): the Amazon “Price Controle” case, decided in February 2026, in which the Bundeskartellamt banned Amazon from using mechanisms that suppressed the visibility of third-party listings it considered too expensive. Yet the underlying proceeding had been opened in October 2020, more than five years earlier, under Article 102 AEUV/Section 19 GWB. And it was the infringement established under those provisions, rather than Section 19a, that formed the basis for the €59 million disgorgement order, corresponding to 1% of the economic benefit Amazon had gained.[4]
Every other case in the group of seven ended somewhere short of a prohibition, activating no obligation under Section 19(a) at all.
Google’s data-processing practices were resolved through commitments in 2023,[5] and its Automotive Services and Maps Platform bundling practices through further commitments in April 2025,[6] in both cases without a finding of unlawful conduct, ad admission of liability, or binding effect on civil courts for follow-on damages claims.
The same pattern emerged in Apple’s App Tracking Transparency Framework (ATTF) case. Following a complaint in May 2021 seeking interim measures, and despite a preliminary finding of an abuse exceeding 300 pages in 2025, the proceeding ended in August 2026 with a commitments decision[7] that complainants consider insufficient to address the concerns.[8] No interim measures, no finding of liability, no fine and no disgorgement.
The Google News Showcase investigation closed in December 2022 after Google promised to abandon some disputed feature – without a formal decision.[9] Meta’s linkage of Oculus/Quest accounts to Facebook logins was resolved the same way in 2024: Meta changed the practice, and the Bundeskartellamt closed the file without using its prohibition power or formally determining what Section 19a would have required.[10] Amazon’s “brand gating” practices, meanwhile, have been under investigation since October 2020.[11]
What others achieved without a Section 19a
Arguably, the more instructive comparison is not to what Section 19a promised, but to what authorities without any equivalent tool have achieved against the same or similar conduct.
Take Google News Showcase. In Germany, the case closed in December 2022 with no decision, after Google agreed to abandon port of the disputed feature. The broader issue did not disappear though. Spain’s CNMC subsequently investigated Google’s News Showcase and closed in December 2025 with fourteen binding commitments covering transparency, disclosure of advertising revenue and impression data, and the methodology for calculating publisher remuneration. The detailed commitments run for five years and may be renewed for another five.[12]
France took an even more determined approach to the same underlying issue (publishers’ compensation for the use of their content) using the general abuse-of-dominance provision rather than anything resembling Section 19a: interim measures against Google in April 2020; a €500 million fine in July 2021 for breaching those measures; binding commitments in June 2022; and a further €250 million fine in March 2024 when Google fell short of them again.[13]
Unlike Germany’s, France’s intervention had an impact. Against the background of its rigorous competition law enforcement, France was the last major Western European market in which Google dared to launch AI Overviews, and compensation of French press publishers exceeds that of their German counterparts by far.
Apple’s App Tracking Transparency framework offers another comparison. The Bundeskartellamt’s ATT proceeding, opened in 2022 before Apple was designated in 2023, closed in August 2026 with commitments.[14] The commitments expressly state that “[n]othing in these Commitments may be construed as implying that Apple agrees with any preliminary views by the Bundeskartellamt .. including on the application and scope of Section 19a GWB/or Art. 102 TFEU”.[15]
France, proceeding entirely under general abuse-of-dominance law, imposed a €150 million fine and adopted a prohibition decision in March 2025, rather than relying on negotiated commitments.[16] Italy’s AGCM independently reached a similar conclusion, fining Apple more €98 million in December 2025 for the same double-consent design, although it had opened its investigation only in 2023.[17] Both proceedings resulted in findings of unlawful conduct and monetary sanctions, facilitating both subsequent negotiations on suitable commitments and damages claims. The German proceeding, despite the availability of Section 19a, produced neither.
Andreas Mundt has emphasised that he is not interested in fines but changing behaviour.[18] This is a legitimate enforcement priority. But it also raises the question whether behavioural change alone provides sufficient deterrence where a proceeding lasts several years and ultimately ends without a finding of infringement or other consequence for the conduct preceding the commitments.
A comparison with Amazon cases points in a similar direction. Italy’s AGCM opened an investigation into Amazon’s marketplace and logistics practices in April 2019[19] and adopted a full infringement decision by December 2021. Finding an abuse of dominance through tying marketplace visibility to use of Amazon’s own logistics service, AGCM imposed a fine of €1.128 billion. A two-year-eight-month turnaround, later substantially upheld on appeal.[20] Germany’s scrutiny of Amazon’s marketplace practices goes back considerably further.[21] Yet its first (and so far only) binding outcome under Section 19a arrived in February 2026, with the prohibition of Amazon’s price-control practices.[22] One authority reached a €1.1 billion infringement decision in less than three years under general abuse rules; the other took more than five years to reach its first Section 19a prohibition, accompanied by a considerably smaller disgorgement order under Article 102 TFEU/Section 19 GWB.
Even beyond the headline gatekeepers, the pattern is similar. When Google refused to allow a competing EV-charging app, Enel X’s JuicePass, to work with Android Auto, Italy’s AGCM opened an investigation in May 2019[23] and fined Google over €100 million by May 2021, two years later.[24] Germany’s structurally similar Google Automotive Services and Maps case took until April 2025 to result in commitments, without a finding of infringement or a monetary sanction.[25]
None of this is to say the French, Italian or Spanish authorities have a perfect enforcement record. Nor are the cases entirely identical. But the comparison does make it difficult to conclude that Section 19a has so far produced the stop-change in speed and effectiveness that accompanied its introduction. Other authorities have reached similar or better outcomes, without a specific provision academics once worried might be too “too powerful”.[26]
Too little, too late?
That gap between ambition and output has not gone unnoticed. Commentators have pointed to the sheer duration of the proceedings as the central problem: Annika Stöhr and Juliane Mendelsohn found that the Bundeskartellamt’s Stage 1 designation decisions grew from already 173 pages for Alphabet to 342 pages for Apple, with the underlying proceedings stretching from eight to twenty-one months. They also concluded that the “easily achieved successes” recorded so far involve conduct already broadly recognised as unfair, while “structural improvements” in the underlying markets remain unlikely.[27] Other commentators argued that the authority is “overwhelmed by the pace of today’s markets”, pointing to cases in which the Bundeskartellamt ended up claiming credit for changes companies made voluntarily.[28]
The government’s own evaluation of Section 19a, which reaches a broadly positive assessment, has been criticises for relying “solely on the basis of the Bundeskartellamt’s own case practice and internal assessments,” without structured input from independent experts or affected market participants.[29] That gap may explain why the draft bill for the twelfth GWB amendment, published in June 2026, contains no reform of Section 19a despite a coalition-agreement’s commitment to further develop German digital competition law to address enforcement deficits.
So far, considerably more institutional effort appears to have gone into designating addresses under Section 19a(1) – and managing tensions with the DMA – than into activating obligations under Section 19a(2), which is where the provision ultimately changes behaviour.
Part of the explanation lies in the inapt structure of Section 19a itself. Because each obligation requires an additional administrative act, while private enforcement is excluded, effective application depends entirely on the Bundeskartellamt. That inevitably creates an enforcement bottleneck. At the same time, a comparison with Brussels, Paris, Rome, and Madrid suggests that institutional design is only part of the answer: the speed and intensity with which the available powers are used also matter.
Making Section 19a Deliver
If Section 19a is to realise more fully the ambition with which it was introduced, three changes merit consideration. The Bundeskartellamt could lower the practical threshold for opening Section 19a(2) investigations and better coordinate designation proceedings with substantive investigations, in particular by ensuring that the same specialised teams handle both. More fundamentally, the legislator should provide that the obligations in Section 19a(2) apply automatically once an undertaking has been designated, akin to the DMA, rather than requiring a separate prohibition decision in each case. That would also allow civil courts to complement public enforcement where administrative resources are necessarily limited. In fast-moving digital markets, where timely relief may matter as much as the ultimate substantive outcome, justice delayed can quickly amount to justice denied.
* Prof Dr Thomas Höppner is partner at Geradin Partners. He was involved in several of the proceedings mentioned in this article.
[2] https://www.competitionpolicyinternational.com/wp-content/uploads/2021/03/6-Section-19a-of-the-Reformed-German-Competition-Act-A-Too-Powerful-Weapon-to-Tame-Big-Tech-By-Jens-Uwe-Franck-Martin-Peitz.pdf
[3] (i) Google Data Processing (ii) Google Maps Platform/Automative Services (iii) Apple ATTF. Note that the Deutscher Bundestag also counts the Google News Showcase investigation as Section 19a proceeding. However, it was triggered by a Section 19 complaint, before Section 19a came into force. https://dserver.bundestag.de/btd/21/054/2105449.pdf
[4] https://www.bundeskartellamt.de/SharedDocs/Meldung/DE/Pressemitteilungen/2026/02_05_2026_Amazon.html
[5] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2023/05_10_2023_Google_Data.html?nn=219148
[6] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2025/04_09_2025_GAS_GMP.html?nn=219148
[7] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2026/08_17_2026_Apple_ATTF.html?nn=55030
[8] https://zaw.de/app/uploads/2026/08/PM_20260818_Apple-ATT_eng.pdf „commitments fall short of the clear antitrust finding.”
[9] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2022/21_12_2022_Google_News_Showcase.html?nn=219148
[10] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2024/10_10_2024_Facebook.html?nn=221094
[11] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2022/14_11_2022_Amazon_19a.html?nn=221082
[12] CNMC, closure decision re: Google/press publishers and news agencies, 17 December 2025, see https://www.uria.com/es/publicaciones/9662-la-cnmc-acepta-los-compromisos-ofrecidos-por-google-y-concluye-su-investigacion
[13] Autorité de la concurrence, decisions 20-MC-01 (9 April 2020, interim measures), 21-D-17 (12 July 2021, €500m fine), 22-D-13 (21 June 2022, commitments) and 24-D-03 (15 March 2024, €250m fine); see press release “Related rights: the Autorité fines Google €250 million for non-compliance with some of its commitments made in June 2022,” https://www.autoritedelaconcurrence.fr/en/press-release/related-rights-autorite-fines-google-eu250-million-non-compliance-some-its
[14] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2026/08_17_2026_Apple_ATTF.html?nn=55030
[15] Case B7-54/25, Commitment Offer pursuant to Section 32b GWB, August 7, 2026 https://www.bundeskartellamt.de/SharedDocs/Publikation/DE/Pressemitteilungen/Apple_ATTF_FAQ.html?nn=52004
[16] https://www.autoritedelaconcurrence.fr/en/press-release/targeted-advertising-autorite-de-la-concurrence-imposes-fine-eu150000000-apple
[17] https://en.agcm.it/en/media/press-releases/2025/12/A561
[18] https://www.youtube.com/watch?v=_NBm6t2ErNQ (Andreas Mundt interviewed by Javier Espinzoa).
[19] https://www.agcm.it/media/comunicati-stampa/2019/4/Amazon-avviata-istruttoria-su-possibile-abuso-di-posizione-dominante-in-marketplace-e-commerce-e-servizi-di-logistica
[20] https://www.agcm.it/media/comunicati-stampa/2021/12/A528-chiusura
[21] Bundeskartellamt scrutiny of Amazon’s marketplace first started in 2010. Investigations were closed (without decision) on 26 November 2013 as well as 17 July 2019.
[22] https://www.bundeskartellamt.de/SharedDocs/Meldung/DE/Pressemitteilungen/2026/02_05_2026_Amazon.html
[23] https://www.tuttoandroid.net/google/agcom-istruttoria-google-app-enel-android-auto-684484/
[24] AGCM, decision A529, “Google/Compatibilità App Enel X Italia con Sistema Android Auto,” 13 May 2021; see also MilanoFinanza, “Google, multa confermata (anche se ridotta) per il caso Enel X,” 2025, https://www.milanofinanza.it/news/google-multa-confermata-anche-se-ridotta-per-il-caso-enel-x-il-no-all-app-juipass-fu-un-abuso-202510291845556883
[25] https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2025/04_09_2025_GAS_GMP.html?nn=219148
[26] https://www.competitionpolicyinternational.com/wp-content/uploads/2021/03/6-Section-19a-of-the-Reformed-German-Competition-Act-A-Too-Powerful-Weapon-to-Tame-Big-Tech-By-Jens-Uwe-Franck-Martin-Peitz.pdf
[27] https://www.tu-ilmenau.de/fileadmin/Bereiche/WM/wth/Diskussionspapier_Nr_184.pdf
[28] https://www.notebookcheck.com/Das-deutsche-Bundeskartellamt-Eine-Behoerde-die-den-eigenen-Anspruechen-hinterherhechelt.733693.0.html
[29] https://www.fiwonline.de/en/bundeswirtschaftsministerium-evaluiert-%C2%A719a-gwb/
Leave a Reply