France’s new legal privilege for in-house counsel: a significant reform with uncertain effects in competition law – August 2026

by Angel Valdes | Aug 25, 2026 | Niet gecategoriseerd

On 23 February 2026, France adopted Law n° 2026-122, introducing a legal privilege for certain legal opinions issued by in-house counsel

The reform acknowledges, for the first time under French law, a distinct legal privilege for in-house counsel alongside the traditional attorney-client privilege reserved for members of the Bar. While this represents a significant improvement for French companies, its practical impact on competition law remains uncertain due to the primacy of European law and the restrictive approach traditionally adopted by French competition authorities and courts.

A new legal privilege subject to strict conditions

The new legal privilege only applies to legal opinions which meet four cumulative conditions:

First, the opinion must be drafted by an in-house lawyer holding a Master’s degree in law (or an equivalent qualification) and having completed a specific ethics training.

Second, it must be submitted exclusively to the management, administrative or supervisory bodies of the employing company, its group or its subsidiaries.

Third, it must constitute legal advice or a legal opinion based on the application or interpretation of a rule of law.

Finally, the document must bear the mandatory wording: “Confidential – Legal Opinion – In-House Counsel.”

Where these conditions are met, the opinion cannot, in principle, be seized or disclosed in civil, commercial or administrative proceedings. However, the privilege is not absolute. It remains subject to judicial review and may be set aside where there are grounds to believe that the document falls outside the statutory requirements or facilitates unlawful conduct.

Most importantly, the new regime is expressly excluded from criminal and tax proceedings and is expressly without prejudice to the investigative powers of EU authorities.

An uncertain impact on competition law proceedings

Whether the reform will significantly strengthen the protection available to companies in competition law proceedings remains uncertain.

At EU level, legal professional privilege remains governed by the AM & S (CJEC, case n°155/79, 18 May 1982) and Akzo Nobel (CJEU, case C-550/07, 14 September 2010) judgments, under which only communications with independent external counsel benefit from protection. The European Commission has recently reaffirmed, in the context of the ongoing review of Regulation n°1/2003, that it sees no justification for extending legal professional privilege to in-house counsel.

French law nevertheless introduces an important distinction. In its decision of 18 February 2026, the French Constitutional Council held that investigations conducted by French authorities under their own statutory powers, including where they apply EU competition law, should not automatically deprive companies of the protection granted by French law. As a result, the new privilege should apply during investigations carried out by the French Competition Authority (“FCA“) in its own name, but not where it acts on behalf of the European Commission.

Although this interpretation significantly enhances the protection available to in-house counsel during national investigations, its practical application remains to be tested.

Practical limitations of the new privilege

Despite this reform, companies should be cautious before relying on the new privilege.

First, the FCA retains the possibility of challenging the confidentiality of legal opinions before the juge des libertés et de la détention (Liberty and Custody Judge, “JLD“), particularly where it considers that a document facilitated or encouraged an infringement of competition law. The Constitutional Council also confirmed that this mechanism applies where the FCA exercises its ordinary powers to request documents.

Secondly, the exclusion of criminal proceedings substantially limits the effectiveness of the reform. As criminal enforcement of competition law is becoming more frequent in France, documents protected in an administrative investigation could ultimately become accessible in the course of subsequent criminal proceedings.

Finally, uncertainty also persists regarding the relationship between the new privilege and attorney-client privilege. French case law continues to adopt a restrictive interpretation of the latter in the context of competition inspections by limiting protection to communications connected with the rights of defense (Cour de cassation, criminal chamber, 13 January 2026, n°24-82.390). This approach appears difficult to reconcile with the broader protection of legal professional privilege recognized by the CJEU and the European Court of Human Rights.

Thomas Lamy
Avocat à la Cour
Grall & Associés

New tools to strengthen competition in Sweden – June 2026

by Angel Valdes | Jun 30, 2026 | Niet gecategoriseerd

The Swedish Parliament has issued a law that will substantially widen and strengthen the administrative tools available to the Competition Authority. The new rules will enter into force on 1 August 2026 and 1 January 2027.

The reform includes new:

(i) substantive provisions to eliminate obstacles to competition,

(ii) possibilities for the Competition Authority to order undertakings to notify future acquisitions below the thresholds

(iii) sanctions for undertakings that provide false data or omit to timely respond to requests for information,

(iv) rules targeting public sector companies that specifically focus on protecting private companies from the unequal conditions that may arise when public and private operate in the same market.

The current Swedish Competition Act upholds prohibitions equivalent to TFEU Art. 101 and 102 as well as rules on control of concentrations. The new rules introduce a possibility for the Competition Authority that now may opt for forward-looking decisions. Hence, the new legislation package allows the Competition Authority to investigate and act against companies in order to remove barriers to effective competition in one or more markets.

The amended merger rules give the Competition Authority a possibility to require an undertaking, for a period not exceeding two years, to report future concentrations to which the undertaking is a party. Such a notification should include details of the other parties involved in the concentration, a description of the concentration, the date of the act forming the basis of the concentration, and the date on which the concentration is intended to be implemented. The new rules apply for concentrations below the thresholds.

Also, there will be sanctions for undertakings that do (i) not provide correct information during an investigation, or (ii) provide information with a delay. The Competition Authority may fine (administrative fine) a company, if it, or anyone acting on its behalf, in the course of an investigation into pro-competitive measures or the assessment of a prohibition on a notified merger, has provided incorrect, incomplete or misleading information in response to a request for information, or has failed to provide the requested information, documents or other materials within the specified time limit.

A new act on public sales activities has also been introduced. The rules in the new act seek to protect private companies from the unequal conditions that arise when public and private undertakings operate in the same market. In addition, publicly owned enterprises will be subject to increased transparency with a new requirement on the contents of their annual reports. These reports should separately state sales activities with (i) a description of the organization and financing of the sales activities; (ii) statement of income and expenses of the sales activities separately from the income and expenses of other activities; and (iii) a specification of the principles and methods applied for the calculation and allocation of income and expenses for different activities.

New legislation will enter into force on 1 August 2026 and 1 January 2027 (transparency requirements).

Cyprus’s Supreme Constitutional Court Upholds €2.1 Million Fine Against Cattle Farmers’ Organisation – May 2026

by Angel Valdes | May 24, 2026 | Niet gecategoriseerd

On 6 April 2026, Cyprus’s Supreme Constitutional Court, in the context of an appeal of the Decision of the Administrative Court, confirmed a €2.1 million fine against the Pancyprian Organisation of Cattle Breeders (POA) that was imposed for several violations of national competition law.

What is this case about?

POA is an organisation that represents cattle farmers across Cyprus. It was found to have abused its position to control how milk was sold and priced, thereby harming fair competition in the market. The Cyprus Commission for the Protection of Competition (CPC) had investigated POA and imposed fines back in 2014.

What did POA actually do wrong?

By way of its decision No. 42/2014, dated 17 October 2014[1], the CPC had established the following violations of Articles 3(1)(a), 3(1)(b) and 6(1)(a) of the Law 13(I)/2008 (equivalent to Articles 101(1)(a), 101(1)(b) and Article 102(a) of the TFEU respectively):

• Violation of Article 3(1)(a) of the Law, regarding practices related to the concluding of exclusive distribution contracts for fresh cow’s milk with cattle-breeders that were members of POA, which included certain standard terms that fixed the purchase price of fresh cow milk. A fine of €600,000 was imposed.
• Violation of Article 3(1)(b) of the Law, regarding the same contracts and the imposition of exclusivity and non-compete clauses for which a fine of €600,000 was imposed.
• Violation of Article 3(1)(b) of the Law, in relation to the ‘Smooth Production of Milk’ (SPM) measure, which essentially tried to control production and for which a fine of €100,000 was imposed.
• Violation of Article 6(1)(a) of the Law, regarding the fixing and/or imposition of excessive and unfair prices by a dominant undertaking. A fine of €800,000 was imposed.

On 24 September 2021, the Administrative Court issued a lengthy 88-page judgment, dismissing POA’s challenge to the lawfulness of the CPC decision.[2]

How did POA try to fight the fines?

POA raised nine separate grounds of appeal before the Supreme Constitutional Court^[3], which were grouped and analysed in three categories:

1. Procedural challenges: Was the process fair?

The Supreme Constitutional Court prioritised these grounds and examined them first since their success would have invalidated the entire proceedings. These procedural grounds of appeal comprised of the argument that the CPC President was biased, that the CPC President had improperly handled her recusal request, that the authority could not fairly act as both an investigator and a judge, and that POA was not given proper access to the case file resulting in the principle of a fair trial not being upheld.

POA relied on newspaper articles and a TV broadcast in which statements attributed to the CPC President allegedly blamed POA before the investigation was concluded. The Supreme Constitutional Court concluded that the alleged statements were neither verified as accurately reported nor proven to be the CPC President’s own words. It additionally found that POA had not discharged its burden of demonstrating bias with sufficient certainty.

The grounds concerning the recusal procedure were also rejected because POA itself had requested an individual presidential decision on recusal and could not then complain that the body had not decided this matter collectively.

As for the challenge to the CPC’s combined role as investigator, prosecutor and judge, this was dismissed by reference to the Filippidou v. Capital Markets Commission (2023) line of authority, which confirmed that Article 6 ECHR is satisfied where, as in this case, a later full judicial review by a competent court is available.

Lastly, on the access to file issue, it was concluded that POA had already had that question decided in an earlier court ruling and could not raise it again now.

2. Substantive challenges: Were the findings correct?

POA essentially tried to argue that its behaviour was mandated, or at least permitted by the national regulatory framework governing recognised agricultural producer organisations. The Supreme Constitutional Court disagreed, finding that POA had pushed its powers to an extreme by forcing members to channel all of their milk exclusively through POA, which went beyond what the framework permitted or encouraged.

POA also attempted to argue that special EU agricultural regulations shielded the agricultural sector from the application of competition law. The Supreme Constitutional Court firmly rejected this, by reference to the well-known decision of the Court of Justice of the European Union (CJEU) in the French endives case[4], which established that producers’ organisations are not automatically exempt from competition law scrutiny.

National competition authorities retain jurisdiction to apply competition rules to agricultural producer organisations and the exemptions from Article 101 TFEU (equivalent to Article 3 of the Law) must be interpreted narrowly. In other words, collective price-fixing and market-sharing agreements are not protected just because an organisation operates in the agricultural sector.

3. The level of the fine: Was €2.1 million too much?

POA challenged the finding that the violations were continuous, which was treated as an aggravating factor in determining the level of the fine that was imposed. The Supreme Constitutional Court disagreed. At no stage did POA make any changes to either its contract terms and exclusivity clauses, nor to the SPM measure or its pricing methodology. It is also pointed out in the judgment that at no point did POA make a statement or show any indication that the said violations had ceased or that remedial measures had been adopted. The finding of continuity was thus found to have been properly justified as an aggravating factor.

Key takeaway

The Cyprus Supreme Constitutional Court upheld every aspect of the original decision. The €2.1 million fine stands and the case is a clear reminder that organisations in the agricultural sector – like any other – must operate within the realm of competition law.

CHRYSSES DEMETRIADES & CO LLC

Ioanna Kyriakidou and Polyvios Panayides

[1]https://www.competition.gov.cy/competition/competition.nsf/All/2E5DCC6D07264C2AC2257EA100400BA1?OpenDocument

[2]https://www.competition.gov.cy/competition/competition.nsf/All/2DE19137F24BDFA8C22587600040C99A?OpenDocument

[3]https://www.competition.gov.cy/competition/competition.nsf/All/388AA9B6AE05A819C2258DD4002440FB?OpenDocument

[4] C-671/15, Président de l’Autorité de la concurrence v. Association des producteurs vendeurs d’endives (APVE) etc.

Paving the way for competitive infrastructure? The CMA’s Civil Engineering Market Study and a £5 Billion Procurement Reform Challenge

by Paul Henty | May 22, 2026 | Niet gecategoriseerd

Paul Henty, Partner and Solicitor

1. Introduction

On 21 May 2026, the UK Competition and Markets Authority published its final report in the Civil Engineering Market Study[1]. After eleven months of evidence-gathering and engagement with governments, procuring authorities and industry, the CMA made nineteen recommendations to the UK government and devolved administrations, seven of which it identifies as critical. The headline opportunity is significant: possible efficiency savings of up to £5 billion per year against estimated annual public expenditure of approximately £19 billion on road and rail infrastructure, excluding HS2. The figure is best understood as an implementation prize, not a forecast saving. It is derived from applying a 10 to 25% efficiency-savings range identified by the National Infrastructure Commission to the CMA’s estimate of annual public road and rail expenditure.

For a European competition audience, two framing points help locate the exercise. A UK market study under the Enterprise Act 2002 is closer in function to a Commission sector inquiry under Article 17 of Regulation 1/2003 than to an enforcement case – it produces findings and recommendations rather than infringement decisions. And although the work is conducted by the competition authority, the substance is overwhelmingly procurement reform rather than antitrust enforcement: supply-side concentration is described as “not inherently concerning,” and the diagnosed problems sit on the demand side.  That distinction matters because the CMA deliberately did not move to a market investigation reference. It found wide-ranging and deep-rooted issues but identified the core concerns as arising from public-policy determined features, including short-term funding settlements and uncertain project pipelines. On that basis, the CMA concluded that formal recommendations were more appropriate and proportionate than the order-making route available after a market investigation reference.

The CMA’s timing is also interesting.  The UK’s new rules on public procurement – the Procurement Act 2023 (PA 23) – took effect only in February 2025.  The CMA did not opt to allow time to see how the new law could affect the competitive landscape.  The report and its findings may be of interest to the EU Commission, which is currently itself in the midst of a review of the EU Public and Utilities Directives.  For the European private sector, the results will also be of interest.  The CMA estimates the value of the UK infrastructure market to be around £19 billion a year in public road and rail expenditure, excluding the High Speed 2 rail project, with the report identifying Siemens and Skanska alongside UK contractors such as Balfour Beatty, Costain, Kier, BAM Nuttall and Murphy.

2. The diagnosis

The CMA finds that the demand side of the infrastructure market in the UK is fragmented: there are six national bodies and hundreds of local authorities making project-by-project decisions in the absence of coordinated strategic direction. Funding cycles are short, pipelines uncertain, procurement practices inconsistent, and regulatory compliance burdensome.

On the supply side, the report estimates that 15 to 20 firms regularly bid for public road and rail contracts, with an average of six bidders per competed National Highways enhancement contract and three to five for contracts in the devolved nations. The CMA’s own caveat is that multiple bidders are “a necessary but not sufficient condition for a competitive tender process” (para 2.26, final report).  Business dynamism (measured by entry and exit rates and the persistence of large firms at the top of the league table) has weakened consistently over the last two decades. SMEs face barriers to entry and scale that the public sector has not effectively addressed.

The findings will read as familiar to anyone tracking procurement debates elsewhere in Europe. What is striking is the CMA’s own acknowledgement, at paragraph 40, of a “persistent failure to track and drive forward the implementation of previous recommendations.” The diagnosis is not the difficult part. The harder question is whether the centre of government can impose durable market-shaping discipline on a sector whose problems arise precisely because decision-making has been fragmented across departments, arm’s-length bodies, devolved administrations and local authorities.  Some of those bodies may be less receptive to additional central direction from the CMA, particularly while still bedding in the PA 23 and related statutory guidance from the Cabinet Office.

3. The recommendation architecture

The nineteen recommendations are organised around five root causes: a fragmented public sector landscape, pipeline uncertainty, capacity constraints, procurement policy and practice, and regulatory barriers. That structure is important because the CMA is not proposing isolated procurement tweaks; it is proposing a market-shaping programme.

• Overarching market shaping (Recommendations 1-2): HM Treasury is asked to take strategic ownership of system-wide reform, supported in practice by the recently established National Infrastructure and Service Transformation Authority. A strategic sector plan should be published with annual progress reporting. This addresses the implementation-failure problem head-on – the question is whether a single department can hold accountability across a landscape this fragmented;
• Addressing pipeline uncertainty (Recommendations 3-5): Multi-year capital budgets of at least three years for all procuring authorities are recommended. Also favoured is greater flexibility for procurers to commit to contracts extending beyond budget settlement periods (the CMA itself flagging that this could favour larger firms but considering the benefit to outweigh). Expansion of the UK-wide infrastructure pipeline to include devolved projects, with funding status and procurement detail visible to the market. The CMA is also realistic about the limits of multi-year funding. Consultation responses warned that even multi-year cycles can be undermined by annual departmental settlements, changing government priorities, poor procurement behaviours and weak pipeline quality. The policy objective is not merely to publish a longer list of projects but to create a credible and funded pipeline capable of changing supplier behaviour;
• Public authority capacity (Recommendations 6-8): Strategic workforce plans for commercial and technical capability. Pooled capacity arrangements so smaller authorities can access specialist expertise. Joint procurement across authorities where appropriate, including more centralised procurement across Network Rail’s five regions. The diagnosis on capacity is acute: the Report records 5,900 civil engineering skills shortage vacancies in 2024. But workforce plans will be judged by whether they address recruitment and retention conditions, not merely whether they reorganise scarce expertise. Without dealing with pay, status and career pathways in the public sector, there is a risk that “capability building” becomes another name for better use of consultants;
• Procurement policy and practice (Recommendations 9-15) This is the largest cluster:
  • Mandatory compliance with the Construction Playbook, ending the “comply or explain” approach.  The Construction Playbook is UK Government guidance, issued by the Cabinet Office, on best practice in the procurement and delivery of public works projects. It sets out core principles on areas such as early market engagement, risk allocation and whole-life value, with the aim of improving consistency and outcomes across public sector construction procurement.
  • Published innovation targets in supply chains every three to five years.
  • All future frameworks to adhere to the Mosey “Gold Standards.” Greater standardisation of procurement processes across authorities.
  • A zero-based review of bespoke “Z clause” contract amendments that distort risk allocation[2].
  • Mandated use of a limited set of standard designs for repeated outputs such as bridges and gantries – the report draws parallels to mandatory national design standards in France, Germany, Canada and Ireland. Strengthened guidance on aligning external designers’ incentives with procurers’.  Several of these are pro-competitive in orthodox terms: process standardisation, reduced bespoke risk amendments, clearer innovation signalling and standard designs should all reduce bid costs, improve comparability and lower incumbency advantages. The caveat is that standardisation must not become a disguised preference for familiar solutions or incumbent delivery models. The Report’s most promising idea is in our view standardisation of repeatable outputs, not standardisation of ambition; and
• Reducing regulatory barriers (Recommendations 16-19). Open challenge functions for industry to challenge out-of-date or duplicative technical standards. A single approved list of supplier accreditations to end overlapping pre-qualification regimes. Fast-track approval processes for innovations in road and rail, including recognition of reference-class data. Standardised utility-diversion response times.

The Report consciously steps back from formal recommendations on planning reform, not because planning is immaterial, but because reforms were already underway. In the main report, the CMA refers to the Planning and Infrastructure Act, proposed updates to the National Planning Policy Framework, statutory consultee reform, DEFRA’s lead environmental regulator model, the Infrastructure (Wales) Act 2024 and Scotland’s NPF4. Appendix A also discusses the Planning and Infrastructure Bill in its account of the reform programme. The CMA therefore focuses its regulatory recommendations on technical standards, accreditations and approvals rather than duplicating the planning reform agenda.

4. A competition-law eye on the package

Most of the package aligns straightforwardly with familiar pro-competitive instincts: lower entry costs, greater transparency, reduced incumbency advantages, faster regulatory throughput. Two recommendations warrant closer attention, because their competition effect is not self-evident from the framing.

The first is Recommendation 4 on longer-term contracts. The Report defines longer-term contracting as contracts spanning more than three years. It presents the dynamic-competition case clearly: longer commitments can justify investment in skills, equipment and innovation, reduce procurement costs and support programme-level learning. The static-competition cost is equally real: the longer the contract, the longer the market is closed to rivals. Indeed, Recommendation 4 may appear to sit oddly with the CMA’s role as a competition regulator. Shorter contracts usually mean more frequent opportunities to compete. But the CMA is looking at dynamic competition, not simply the number of tender events. If suppliers are expected to invest in people, equipment, technology and innovation, they need a credible opportunity to earn a return on that investment. As any engineering consultancy will recognise, participating in complex infrastructure procurements is itself expensive and time-consuming. A market with frequent tenders but weak investment incentives may be contestable in form yet underpowered in substance.

Article 33 of Directive 2014/24/EU caps framework agreements at four years for precisely this reason. That EU comparison is not directly applicable to all UK arrangements post-Brexit, but it remains analytically useful because it captures the same foreclosure concern.

The missing operational link is to the Procurement Act 2023 (PA 23) architecture. Open frameworks (s. 49 PA 23) and dynamic markets (s 34-40 PA 23) are designed, in different ways, to keep opportunities contestable over time. Open frameworks, a framework agreement with a break point to allow for new competition, are a UK innovation which EU procurement rules do not yet have.  A fuller Recommendation 4 would have explained when longer contracts should be channelled through those mechanisms, when break clauses or periodic re-tendering should be expected, and when direct long-duration commitments are justified by safety-critical continuity or programme-level efficiencies. The Report also records Network Rail’s evidence that it avoids break clauses because they are “potentially costly and therefore riskier,” which suggests the cultural barrier to mitigation is itself part of the problem.

The second issue sits at paragraph 22 of the Executive Summary, on prior experience. The CMA correctly identifies that requirements for specific prior experience or a proven track record can favour incumbents even where firms have relevant capability from adjacent markets. That is a classic procurement competition problem: the authority wants assurance, but the chosen proxy narrows the market. The Report records a useful example of what reform looks like in practice – Network Rail told the CMA it now focuses on broader infrastructure experience rather than rail-specific experience, to widen the addressable supplier base. The Report would have been stronger had it converted that insight into a defensible procurement template: acceptable capability proxies, treatment of transferable experience, worked examples of weighting, and model drafting capable of withstanding bidder challenge.

5. What to watch

The CMA has produced a careful diagnostic and a coherent package. Several recommendations should be relatively deliverable, particularly those aimed at lowering regulatory barriers, simplifying accreditation, standardising procurement processes and reducing unnecessary bespoke risk allocation. The harder recommendations are those that require government to change its own behaviour: multi-year funding discipline, credible pipelines, public-sector capability and genuine strategic ownership.

Now that the Report has been published, the UK Government has committed to respond within 90 days.  Its response will likely set out the extent to which it accepts the CMA’s findings, along with the steps it proposes to adopt to implement these.

The Report’s most important sentence may therefore be its acknowledgement of the “persistent failure to track and drive forward the implementation of previous recommendations.” The £5 billion opportunity is not in the analysis. It lies in whether government can make the procurement system behave differently, because in infrastructure, as in procurement, old habits die hard.

[1] Civil engineering market study – GOV.UK

[2] “Z clauses” are bespoke amendments to standard form contracts (most commonly NEC contracts) used in UK construction procurement. While they can address project-specific requirements, they are often criticised where they materially alter the standard allocation of risk, add complexity, or reintroduce issues that standard forms are designed to avoid.

German Federal Cartel Office Prohibits Amazon’s Price-Control Mechanisms: A Landmark Decision on Hybrid Platforms – April 2026

by Angel Valdes | Apr 22, 2026 | Niet gecategoriseerd

On 5 February 2026, Germany’s Federal Cartel Office (Bundeskartellamt, “FCO”) announced a far-reaching prohibition directed at Amazon.com, Inc. (Seattle, USA) and Amazon EU S.à r.l. (Luxembourg) (together “Amazon”). The authority bars Amazon from using mechanisms that influence the prices set by third‑party sellers on the German Amazon Marketplace. Such price-control mechanisms may in future be deployed only in exceptional situations of excessive pricing—and then only under strict parameters and transparency requirements specified by the FCO.

The decision is notable not only for its substantive outcome—curtailing a major platform operator’s levers over seller pricing and visibility—but also for its legal basis and remedial architecture. The FCO relies on the special regime for major digital companies under section 19a(2) of the German Competition Act (Gesetz gegen Wettbewerbsbeschränkungen, “GWB”), alongside the general dominance provisions of section 19 GWB and Article 102 TFEU. In addition, the authority, for the first time, orders disgorgement of the economic benefit allegedly obtained from the infringement, fixing an initial partial amount of approximately EUR 59 million.

The FCO’s intervention sends a clear signal on how Germany intends to police conflicts of interest inherent to “hybrid platforms”, and how transparency obligations towards business users may be woven into competition law enforcement.

1. Amazon’s Marketplace as a Hybrid Platform and Its Market Position

Amazon operates a broad digital e‑commerce ecosystem. In Germany, amazon.de is not merely an online shop but also a marketplace that enables third‑party sellers (“marketplace sellers”) to sell goods directly to end customers. Amazon is present in two roles on the same user interface: first, as a retailer through its own trading business (“Amazon Retail”); and second, as the operator of the marketplace infrastructure (“Amazon Marketplace”), providing access, listing, and transaction-related services to third‑party sellers in exchange for fees and commissions.

This dual role is at the heart of the competition law concerns. In competition policy terminology, a platform on which the operator both sets the rules of display and ranking and simultaneously competes downstream with the business users is referred to as a “hybrid platform”. The FCO emphasises that this structural feature carries heightened risk, especially where the platform’s market significance makes it difficult for sellers to forego access.

According to the FCO, more than 60% of German online retail turnover in goods is generated via Amazon’s trading platform. The site reportedly lists around 1.5 billion distinct items, with more than 200,000 third‑party sellers active on the marketplace. Third‑party sellers account for approximately 60% of the total trading volume on amazon.de, while Amazon Retail accounts for roughly 40%. In the specific market for marketplace services provided to commercial sellers in Germany, the authority points to a turnover-based market share exceeding 70%.

This market context matters because the measures at issue do not simply affect a seller’s price point in isolation. Rather, they operate through the platform’s most valuable asset: visibility and access to demand. Restrictions on visibility can rapidly translate into severe revenue losses—raising the risk of foreclosure of (often small and medium-sized) sellers and, in the authority’s view, distorting the competitive process both within the marketplace and vis‑à‑vis other e‑commerce channels outside Amazon.

2. The Role of the “Buy Box” and the Economic Meaning of Visibility

A central point in the FCO’s case is the “Buy Box” (referred to in Amazon’s more recent terminology as the “Featured Offer” in the purchase field). On a marketplace with many sellers listing identical or substitutable products, the platform must organise presentation through ranking criteria; otherwise customers would struggle to locate relevant offers. On amazon.de, customers typically search products via keywords, receive lists of items, and then click through to product detail pages. These detail pages usually feature a prominently highlighted purchase area—the Buy Box—displaying one offer selected by Amazon along with price and shipping information, coupled with the “Add to Cart” and “Buy Now” buttons.

The selection of the offer displayed in the Buy Box is determined by an algorithm applying multiple criteria—commonly including price, delivery speed, and other performance indicators. The FCO underlines that the overwhelming share of the platform’s trading volume is realised via the Buy Box. Consequently, exclusion from Buy Box consideration is not a neutral, technical re-ranking; it is a major commercial disadvantage that can push sellers into subordinate placement (e.g., the “Other sellers on Amazon” section) from which only a small portion of sales is typically generated. In some cases, where all offers are disqualified, the product page may show only an “All offers” field rather than a featured merchant offer.

Against this background, any mechanism that ties Buy Box eligibility to Amazon-defined price thresholds can effectively function as a de facto price cap—especially for sellers who depend economically on Amazon as a route to customers.

3. The Price-Control Mechanisms Challenged by the FCO

The FCO’s decision targets Amazon’s use of various “price-control mechanisms” applied to third‑party seller offers. These mechanisms are anchored, according to the investigation, in Amazon’s terms and policies, including the “Marketplace Fair Pricing Policy”. If the mechanisms categorise an offer as too expensive, Amazon either removes it from the marketplace entirely or limits its visibility—most importantly by preventing the offer from being displayed as the Featured Offer/Buy Box.

The authority describes three principal mechanisms:

• “Price error” mechanism leading to deactivation.

The first category concerns prices characterised by Amazon as “price errors”. Using a statistical model developed by Amazon, offers deemed excessively high and therefore potentially erroneous are removed from the marketplace (“deactivated”).

• “Too high price” mechanism leading to Buy Box disqualification.

A second mechanism, also based on a statistical model, categorises prices as “too high”. Rather than removing the offer entirely, Amazon disqualifies it from Buy Box consideration, thereby pushing it into less visible areas of the product page.

• “Not competitive price” mechanism based on external price comparisons.

A third mechanism is applied where products are also offered in other online shops or on other marketplaces. Amazon performs ongoing comparisons with a set of online retailers maintained on an internal list. Based on this, Amazon identifies a “competitive price” corresponding to the lowest currently observed price among those retailers. The FCO notes that Amazon may disregard the shipping costs charged by the external retailer, while requiring the marketplace seller’s price including shipping not to exceed the computed threshold to avoid Buy Box disqualification.

The FCO criticises not only the impact of these mechanisms but also their opacity. In its view, third‑party sellers are not sufficiently informed about how the relevant price ceilings are derived, where those ceilings approximately lie, and under what concrete circumstances their offer will be removed or become only partially visible. The authority further states that Amazon does not disclose—either in its contractual documentation or in explanatory materials—how it decides whether a price triggers deactivation rather than mere disqualification, nor the broad functioning of the statistical models generating these thresholds.

From a practical standpoint, these mechanisms can force sellers to adjust prices to avoid losing visibility. Because sellers bear their own economic risk and are responsible for setting prices, the authority sees this practice as a systematic interference with sellers’ pricing freedom, capable of preventing sellers from covering costs and, in the extreme, leading to their displacement from the platform.

4. Competition Law Assessment under § 19a GWB, § 19 GWB and Article 102 TFEU

The FCO qualifies Amazon’s conduct as an abuse under multiple legal bases.

First, the authority relies on section 19a(2) GWB, the special framework introduced to address large digital ecosystems with “paramount significance across markets”. The FCO had already determined in July 2022 that Amazon meets this threshold. The German Federal Court of Justice confirmed that designation in April 2024. Section 19a GWB is designed to enable faster and more targeted intervention against practices by digital conglomerates that may harm competition even beyond a single narrowly defined market.

Second, the FCO also invokes the general dominance provisions under section 19 GWB and Article 102 TFEU. The authority considers Amazon to hold a dominant position in Germany in the market for marketplace services to commercial sellers, supporting this by reference to shares exceeding 70% (turnover-based) and a leading position by user numbers.

Substantively, the FCO’s concern is that Amazon’s intervention in sellers’ pricing, implemented via visibility sanctions and Buy Box exclusion, can distort the competitive process on the marketplace. Because Amazon competes directly with sellers on the same platform, imposing price ceilings on those sellers—even in the form of de facto caps linked to Buy Box eligibility—raises the risk of Amazon steering the platform-wide price level according to its own preferences. The authority links this to two main competitive harms:

• Restriction and coordination of competition within the marketplace.

By applying frequently changing price thresholds set at Amazon’s discretion and not grounded in objective and verifiable principles, the platform can shape the competitive process between sellers and limit sellers’ ability to pursue independent pricing strategies.

• Foreclosure and reduced contestability beyond Amazon.

The FCO warns that enforcing the lowest observed external online price within Amazon may deter other online retailers outside Amazon from “price attacks”. If any external price cut is rapidly mirrored on Amazon through the marketplace mechanism, the external retailer cannot attract a meaningful customer base with the lower price, reducing the incentive to offer better deals outside Amazon and increasing customer lock‑in.

Importantly, the FCO states that it does not oppose Amazon’s goal of providing low prices to consumers. Rather, it argues that the chosen means—punishing or suppressing seller offers because their prices exceed Amazon-defined thresholds—are not necessary to pursue that goal and are competition‑restrictive. The authority points to less harmful alternatives, such as lowering fees and commissions charged to sellers, thereby creating incentives for sellers to pass on cost reductions to consumers.

5. Remedies and Forward-Looking Conditions: Exceptional Use Only, With Strict Transparency

The decision prohibits Amazon from applying the existing mechanisms in their current form. Looking forward, the FCO does not categorically ban any and all intervention in seller pricing. Instead, it sets a narrow corridor: price-control tools may be used only exceptionally, particularly where a seller’s pricing violates applicable law—explicitly mentioning usurious pricing and “usury-like” price conditions. If Amazon seeks to use a price-control mechanism within this narrow scope, the FCO demands significantly enhanced transparency and procedural clarity.

These obligations intersect with broader platform regulation. The FCO notes that transparency requirements also need to comply with the EU Platform-to-Business Regulation (P2B Regulation), which governs fairness and transparency in platform relationships with business users. In Germany, enforcement of the P2B Regulation lies with the Federal Network Agency (Bundesnetzagentur). The FCO states that it coordinated its transparency-related aspects with that authority.

6. Disgorgement of Economic Benefit: An Emerging Remedy in Digital Antitrust Enforcement

A particularly noteworthy aspect is the FCO’s resort to disgorgement of the economic benefit gained through the infringement. The authority is applying, for the first time, the revised disgorgement instrument that was introduced in 2023. Under the new regime, the economic advantage can be established via a presumption, reducing the evidentiary burden and facilitating practical application.

In this case, because the FCO views the infringement as ongoing, it initially set only a partial amount—approximately EUR 59 million. The authority stresses that disgorgement is not a fine intended to punish; rather, it aims to neutralise the benefit and ensure that the infringing company cannot retain gains derived from unlawful conduct. It is not entirely clear how the FCO estimated the economic benefit resulting from the price control mechanism. There are basically two possible ways how Amazon could have benefitted from the illicit practices: if the price control mechanism led to higher prices on the platform, Amazon would have benefited from this through its own sales via Amazon Retail. If the price control mechanism led to lower prices, this would have attracted additional traffic to the detriment of other e-commerce platforms. Amazon would then have benefited from a higher volume of commissions paid for sales on its platform.

If upheld and applied more broadly, this tool could become a significant complement to behavioural remedies in digital markets, where conduct remedies may take time to implement and where structural economic gains may accrue during litigation.

7. Outlook: Implications for Platform Governance and Seller Autonomy

The Amazon decision illustrates a sharpened stance on how competition authorities view platform levers that indirectly yet powerfully shape market outcomes. Two themes stand out.

First, the case underscores that on a dominant hybrid platform, “visibility governance” can be functionally equivalent to direct price regulation. Excluding offers from the Buy Box or removing them from the marketplace because they exceed algorithmically determined thresholds can pressure sellers into adopting platform‑preferred pricing, even where the platform claims a consumer‑welfare justification.

Second, transparency is increasingly treated not merely as a matter of consumer information or contract fairness, but as a competition parameter. Where a platform’s internal models and thresholds are opaque, sellers cannot plan, challenge, or adapt effectively, and competitive constraints may be weakened. The FCO’s detailed requirements on how mechanisms must be described, updated, and communicated point to a more proceduralised conception of competition compliance in platform settings.

Moritz Lorenz
Berlin