Turkish competition board fines seventeen pharmaceutical companies for no-poach agreements and wage data exchange

1. Introduction

On 8 April 2026, the Turkish Competition Board (“Board”) published its decision dated 11 September 2025 and numbered 25-34/810-474 concluding a broad multi-party investigation into anti-competitive conduct in the pharmaceutical sector. The allegations were twofold: that pharmaceutical companies had concluded no-poach agreements and exchanged competitively sensitive information in the labour market, and that AbbVie had separately abused a dominant position in the Hepatitis C antiviral treatments market through unlawful payments to physicians, donation of medical devices to divert prescribing preferences, obstruction of competitors’ market entry via manipulation of licensing and reimbursement processes, and attempts to obtain confidential data from a competitor’s Early Access Programme. Of the 30 undertakings investigated, 17 were found to have infringed Article 4 of Law No. 4054 on the Protection of Competition (“Law No. 4054”), with total administrative fines of approximately TRY 244.8 million. The decision provides important guidance on fine calculations under the new penalty regulation, the ancillary restraint doctrine, and information exchange analysis in the HR context.

2. Labour market violations: no-poach agreements and information exchange

2.1 No-poach agreements

The Board uncovered systematic practices by which certain undertakings refrained from hiring employees from designated competitors. Those companies were labelled internally as “off-limits,” “restricted” or subject to a “gentlemen’s agreement,” and in several recruitment processes specific firms were excluded from the candidate pool at the outset, with internal documents explicitly noting that candidates were disqualified on that basis. The Board confirmed that such arrangements constitute infringements of Article 4 of Law No. 4054 by object, regardless of any actual effect. It characterised no-poach arrangements as labour input allocation agreements equivalent to customer or market allocation cartels, suppressing wages and depriving employees of access to better opportunities. Sanovel had the longest and most extensive no-poach conduct on file, with evidence spanning March 2013 to March 2023 (nearly ten years) and involving six counterparties: Abdi Ibrahim, Bilim, Berko, Genveon, Ilko, and Santa Farma. A separate bilateral information exchange infringement was also found with Menarini. Duration directly drove the highest fine: Sanovel received TRY 79,358,342, calculated on a base rate with a 100% duration uplift for infringements exceeding five years.

2.2 The “mutual hiring” defence

Certain undertakings argued that actual employee transfers during the relevant period negated any no-poach understanding. The Board rejected this defence. Since no-poach arrangements are restrictions by object, demonstration of implementation or anti-competitive effects is unnecessary; transfers remain probable even where such an arrangement exists. The relevant question is whether employee mobility was restricted, and documentary evidence of intent to restrict hiring suffices to sustain the infringement finding.

2.3. Ancillary restraints

The Board assessed no-poach and non-solicitation clauses embedded in commercial agreements under the ancillary restraint doctrine, applying a two-limbed test: the restriction must be directly related to, and objectively necessary for, the implementation of the principal transaction. It further assessed proportionality by reference to duration, personal scope, and subject-matter scope. The logistical distribution agreement (“LDS”) between AbbVie and Abdi Ibrahim included a mutual non-solicitation clause extending twelve months beyond termination. The Board had previously reviewed the LDS in an individual exemption decision dated 17 June 2010 and numbered 10-44/784-261, refusing exemption on the non-compete clause but raising no objection to the non-solicitation clause. The Board treated this silence as implicit acceptance of the non-solicitation clause as an ancillary restraint and considered that earlier position relevant in the present case. In the licence and co-promotion agreement between Menarini Asia Pacific Holdings and AstraZeneca UK, the Board identified a non-solicitation clause preventing the recruitment of AstraZeneca group employees for sixty months. It accepted the clause as an ancillary restraint, noting that the restriction was tied to the transfer of product-specific know-how, limited to exposed personnel, and confined to a defined duration. However, the Board’s primary finding was that the arrangement had no direct, foreseeable, or significant effects on Turkish markets and therefore fell outside Law No. 4054 under the effects doctrine in Article 2. The ancillary restraint analysis was only alternative, so the accepted duration should be treated with caution. The Board also accepted non-solicitation clauses in confidentiality agreements linked to pre-acquisition due diligence. In one case, a twelve-month restriction was accepted as necessary to protect sensitive information exchanged during a potential transaction, limited to employees in Germany identified through that process. Again, the Board’s primary finding was that the agreement fell outside Article 2 under the effects doctrine, and the ancillary analysis was secondary. A different situation arose in the Bausch & Lomb–Liba relationship. An internal WhatsApp message dated March 2021 referred to a “gentlemen’s agreement” suggesting that Bausch & Lomb refrained from recruiting Liba’s employees during a distribution agreement concluded in 2016 between their foreign affiliated companies, covering marketing authorisation, distribution, and promotion in Turkey. Although not a signatory, Bausch & Lomb was found to be part of the same economic unit and thus an indirect party. The Board treated the practice as a possible ancillary restraint, finding a direct link based on the nature of the distribution relationship even without an express clause. On necessity, it noted that Liba’s workforce had fallen from approximately 104–106 employees in 2019–2020 to around 20 by 2024, with employees working across functions such that losing even one could affect performance. The Board also accepted that the relationship was vertical, not horizontal, and assessed the conduct within that framework rather than as a standalone no-poach agreement. It further noted that Bausch & Lomb had hired Liba employees and that there was no evidence Liba was prevented from doing the same. On this basis, the Board found no infringement of Article 4.

2.4 Multilateral information exchange in the labour market

A separate strand of the investigation concerned a multilateral exchange of salary-related data. The central piece of evidence, dated 11 July 2023, was an email circulated among HR managers of eight companies. It attached a file containing each company’s ordinary and supplementary wage increase rates, planned additional increase percentages, and meal allowance data, including forward-looking entries. Amgen, AstraZeneca, GSK, Merck, Novartis, Novo Nordisk, Pfizer, and Sanofi all participated. GSK, although identified as a participant, does not appear in the fine list, having settled prior to the final decision. A separate bilateral exchange between Sanovel and Menarini, involving 2022 wage increase rates shared at general manager level, was also found to constitute an infringement. The Board confirmed that both exchanges constitute Article 4 infringements by object, consistent with the Horizontal Cooperation Guidelines’ treatment of forward-looking HR data as competitively sensitive. Paragraph 43 of those Guidelines is expressly cited, noting that information sharing enables competitors to gain insights into each other’s market strategies and facilitates the alignment of pricing policies.

3. AbbVie: no infringement found under articles 4 or 6

3.1 Article 4 and the succession principle

AbbVie was alleged to have concluded no-poach arrangements and exchanged competitively sensitive information with various counterparties. A threshold question arose because key evidence referenced Allergan, over which AbbVie subsequently obtained sole control. The Board confirmed that under Turkish law, universal succession governs acquisitions, and the acquiring undertaking assumes responsibility for the acquired entity’s competition law infringements. Despite this, the Board found no infringement under Article 4. On the no-poach allegations, evidence that employee transfers had continued throughout the period under review, undercutting any effective no-poach understanding. On the information exchange allegations, the Board concluded either that the information in question (e.g., Turkish Social Security Institution (“SSI”) – published discount rates) was not sufficiently strategic or forward-looking to reduce market uncertainty, or that AbbVie was not a recipient of the relevant email chain and no direct inter-company communication regarding the multilateral wage exchange was on file.

3.2 Article 6: dominance and alleged abuse in the hepatitis c market

The Board defined the relevant market as Hepatitis C antiviral treatments at ATC-5 level, found it to constitute an effective duopoly between AbbVie (Maviret/Viekirax-Exviera) and Gilead (Harvoni/Sovaldi/Vosevi), and concluded that AbbVie held a dominant position for 2018–2024. The countervailing buyer power of the SSI did not suffice to negate dominance. Notwithstanding the dominance finding, the Board concluded that none of the three categories of alleged abuse was established. Honoraria and consultancy payments fell within regulatory limits and reflected legitimate commercial or scientific activity. The Alternative Reimbursement System brought both AbbVie’s and Gilead’s products within scope simultaneously, internal messages reflected dissatisfaction with reimbursement terms rather than unlawful pressure, and a related episode showed that AbbVie did not always obtain its preferred regulatory outcome. The patient data estimation mechanism aggregated AbbVie’s own estimated patient numbers without extracting competitor data, competitor activity updates constituted ordinary field intelligence, and CRM data was collected via physician surveys without capturing personal patient data.

4. Administrative fines: key methodology points

The Board applied the more favourable of the repealed 2009 Regulation and the new Regulation (in force since 27 December 2024) on a company-by-company basis, creating complexity where co-infringers are subject to different regimes in the same decision. In practice, only Adeka was fined under the old Regulation. Under the old Regulation, no-poach arrangements attracted the higher cartel base rate of 2–4% of annual gross revenue. Under the new Regulation, the binary cartel/other infringement split is abolished, but such conduct remains characterised as a serious infringement. Wage and benefit data exchanges, classified as restrictions by object outside the cartel bracket, attracted a lower base rate—a distinction with direct financial consequences. The Board did not automatically treat the period between the first and last piece of evidence as a continuous infringement; where evidentiary gaps were significant, duration was reduced. For Argis, evidence dated from July 2018 to June 2022, but a gap exceeding three years reduced the effective infringement duration to less than one year. Several companies successfully argued that revenues generated outside Turkey should be treated as a mitigating factor, given that the conduct related solely to the domestic labour market. Combined with the removal of cartel-specific brackets under the new Regulation, this consistently produced more favourable outcomes for companies with significant export revenues. A Board Member dissented on fine calculation methodology, arguing with reference to Council of State authority that the fine base should be net sales, on the ground that the relevant provisions define annual gross revenue by reference to net sales in the standard chart of accounts. In his view, calculating fines by reference to employee costs as a proportion of total turnover is inconsistent with both the statutory text and settled judicial interpretation. Practitioners advising on challenges to fine quantum should note this disagreement.

5. Conclusion

This decision is one of the most extensive labour market enforcement actions in Turkey to date, confirming that no-poach arrangements are cartel-equivalent restrictions by object and extending the same logic to HR information exchanges. The breadth of the investigation, its duration, and the total fines imposed signal that labour market enforcement is a primary concern under Turkish competition law. The decision advances practice on three fronts. First, the rejection of the mutual hiring defence is unequivocal: documentary evidence of intent to restrict hiring suffices, regardless of whether transfers continued. Second, the calibrated application of the ancillary restraint framework across transaction types provides detailed guidance, distinguishing M&A scenarios from ongoing commercial arrangements. Third, the more-favourable-law analysis, applied company by company, illustrates the practical complexity when a new penalty regulation enters into force during an ongoing multi-party investigation. For practitioners, the compliance implications are immediate. Categorical restrictions on hiring from named competitors, whether formal or informal, constitute per se Article 4 infringements. Salary data and benefit parameters must not be shared with competitors, even informally. Non-solicitation clauses in commercial agreements should be narrowly drafted, confined to identified personnel, proportionate in duration, and directly linked to the principal transaction’s objective. Revenues attributable to activities outside Turkey should be documented and presented as a mitigating factor in penalty proceedings. Finally, acquirers assume competition law liability for acquired entities under universal succession, making pre-closing review of labour market practices essential to transaction risk assessment.

Originally published in Concurrences