ICLE Comments to UK ‘Refining Our Competition Regime’ Consultation
Executive Summary
The International Center for Law & Economics (‘ICLE’) respectfully submits these comments in response to the Department for Business and Trade’s consultation on Refining Our Competition Regime.[1] ICLE is a nonprofit, nonpartisan research centre that promotes rigorous economic analysis in public-policy debates.
ICLE welcomes the Government’s commitment to economic growth and its recognition that pace, predictability, proportionality, and process—the CMA’s ‘4Ps’ framework[2]—are essential to maintaining the United Kingdom as an attractive jurisdiction for investment and innovation. Several proposals in the consultation reflect sound, evidence-based policymaking. Others, particularly the most far-reaching structural reforms, risk undermining those same objectives.
A central concern is that the consultation adopts, and in some cases generalises, institutional features drawn from the Digital Markets, Competition and Consumers Act 2024 (DMCC Act). That regime is recent, narrowly targeted, and largely untested in practice. Extending its most intrusive features across the full scope of competition and consumer enforcement risks importing uncertainty into areas where predictability is critical. Reform should proceed incrementally, not by transplanting a sector-specific regulatory model to the wider economy.
Measured against the Government’s own ‘4Ps’ framework, the Strategic Steer’s emphasis on independence and proportionality,[3] and established law & economics principles—including the error-cost framework and the consumer-welfare standard—ICLE’s principal concerns and recommendations are as follows:
- The proposed abolition of the CMA’s independent panel for Phase 2 merger and market investigations should not proceed without the simultaneous introduction of merits-based appellate review at the Competition Appeal Tribunal. The panel provides a quasi-judicial ‘fresh pair of eyes’ and underpins the current judicial-review standard. Removing it without compensating safeguards would concentrate investigative, prosecutorial, and adjudicative authority within a single institution, increasing the risk of confirmation bias and weakening procedural fairness.
- Extending algorithmic investigation powers from the DMCC Act’s Strategic Market Status regime to general competition enforcement would be internationally unprecedented and disproportionate to demonstrated need. It would subject firms across the economy to quasi-regulatory oversight designed for a narrow, ex ante framework, raise serious trade-secret and confidentiality risks, and impose burdens that may deter innovation and investment.
- The statutory codification of sunset clauses for market remedies is welcome and should be strengthened. A risk-based review cadence, clearer petition mechanisms with defined evidentiary standards and binding response timelines, and consistent transparency requirements would ensure that remedies remain proportionate as markets evolve.
- Expanding concurrency by allowing sector regulators to oversee CMA-imposed remedies would risk duplicative oversight, fragmented accountability, and higher compliance costs. The CMA should remain the single lead authority for remedy oversight, with sector regulators contributing expertise through a consultative model.
- More broadly, reforms should preserve clear institutional boundaries and accountability. Concentrating decision-making authority, expanding investigatory reach, and fragmenting oversight across multiple regulators would collectively reduce predictability and increase enforcement error.
- All reforms should be guided by the error-cost framework. Because false positives in competition enforcement are generally more costly and less self-correcting than false negatives—particularly in innovation-intensive markets—expanding CMA powers requires stronger evidentiary standards, more robust appellate review, and effective sunset mechanisms.
In sum, the Government’s ambition to create a ‘best-in-class’ competition regime calls for a selective and disciplined approach. Targeted procedural improvements should be prioritised. Reforms that expand powers without commensurate safeguards risk increasing uncertainty, deterring investment, and undermining the pro-growth objectives at the centre of the Government’s agenda.
I. Political Reorientation of UK Competition Policy
The consultation follows an 18-month period of unusually direct political intervention in UK competition policy. The Labour government, elected in July 2024, made economic growth its ‘number one mission’.[4] Prime Minister Keir Starmer compared UK regulation to ‘Japanese knotweed’, while Chancellor Rachel Reeves told regulators in January 2025 that ‘every regulator, no matter what sector, has a part to play by tearing down the regulatory barriers that hold back growth’.[5]
The CMA quickly became the focal point of this dissatisfaction. Its April 2023 decision to block the $68.7 billion Microsoft/Activision Blizzard merger—later reversed after restructuring—came to symbolise a regime seen as prioritising jurisdictional reach over growth and investment confidence. On 21 January 2025, Business Secretary Jonathan Reynolds removed CMA Chair Marcus Bokkerink. In his farewell remarks, Bokkerink warned that ‘investors put a price on the risk of political intervention, unpredictability and inconsistency’.[6] Reynolds replaced him with Doug Gurr, former head of Amazon UK, and made clear that the Government expected regulators to ‘supercharg[e] the economy’.[7]
The Strategic Steer to the CMA, published 15 May 2025, places economic growth at the centre of the authority’s mandate. It directs the CMA to prioritise pro-growth, pro-investment interventions, focus on harms with a ‘clear and direct UK impact’, support eight Industrial Strategy growth sectors, and consider whether other jurisdictions could address UK concerns through parallel enforcement.[8] The CMA operationalised this shift through its ‘4Ps’ framework—Pace, Predictability, Proportionality, and Process—introduced in late 2024 and formalised in February 2025.[9]
Institutional changes followed quickly. By March 2025, the CMA had issued a Mergers Charter, launched a merger-remedies review, and adopted new Phase 1 performance targets, including a 40-working-day pre-notification period and 25-working-day clearance for straightforward cases.[10] In October 2025, the Government reinforced this direction by introducing a stronger statutory growth duty and a public accountability dashboard.[11] The CMA conducted no Phase 2 merger investigations in the year preceding the consultation—a striking indication of the behavioural shift already underway.
Against this backdrop, a central concern in these comments is that several of the consultation’s most consequential proposals draw heavily on the Digital Markets, Competition and Consumers Act 2024 (DMCC Act). The DMCC’s digital-markets regime remains largely untested. It has operated for barely a year and applies to only a small number of cases. Extending this framework across the CMA’s broader competition and consumer functions risks importing uncertainty into areas where predictability remains essential.
That risk sits uneasily with the Strategic Steer itself, which instructs the CMA to apply the DMCC regime ‘flexibly, proportionately and collaboratively’. Expanding its most intrusive features before they have been tested in their original context is unlikely to advance those objectives. It instead risks compounding the very concerns about proportionality and predictability that the steer seeks to address.
The Government’s consultation recognises that the reforms aim to improve ‘the pace, predictability, proportionality, and process of engagement through refinements to the legislative framework for competition’.[12] Several proposals move in the opposite direction. In particular, the single-phase market review tool, the expansion of algorithmic investigation powers, and the removal of the independent panel without adequate safeguards would reduce predictability and weaken proportionality. These changes would undercut, rather than support, the Government’s stated objectives.
II. Abolition of the Panel System and Its Implication
The Government’s proposal to abolish the panel-led Inquiry Group system and replace it with Board-appointed sub-committees marks the most consequential structural reform in the consultation. It is also unlikely to advance the Government’s stated objectives.
The panel system has deep institutional roots. It traces back to the Monopolies and Restrictive Practices Commission established under the Monopolies and Restrictive Practices (Inquiries and Control) Act 1948. The CMA retained this model to ensure independent, quasi-judicial decision-making in complex competition cases. Inquiry Groups, drawn from a panel of approximately 30 members appointed by the Secretary of State, bring ‘fresh eyes’ to each case.[13] They operate under a chair appointed by the CMA Chair with Board consent, and they function as an independent decision-making body within the CMA’s broader institutional framework.
The proposal would dismantle this structure. Decisions currently reserved to Inquiry Groups would shift to the CMA Board, to new Board-level committees (including a Merger Board Committee and a Markets Board Committee), or to sub-committees appointed by those bodies.[14] The Board could also delegate certain decisions to case teams, subject to internal governance procedures.[15]
The Government justifies this shift on grounds of accountability, predictability, and consistency. The CMA Board answers to Parliament, while panel members do not. At present, the Chair, CEO, and Board remain accountable for outcomes in which they cannot participate directly. The Government argues that a Board-led model would resolve this perceived mismatch and produce more consistent decision-making.[16]
That argument is unconvincing. The current framework does not create a meaningful accountability gap.[17] The Board already plays a central role in Phase 2 processes. It initiates and oversees market studies, makes market investigation references, and issues advisory steers that Inquiry Groups must consider. The Board therefore shapes the direction and scope of investigations, even if it does not take final decisions. That division is deliberate. It preserves a structural separation between oversight and adjudication.
This separation reflects a conscious legislative choice, not an institutional flaw. Parliament designed the system to insulate case-specific decisions from a Board that is itself subject to government direction. The accountability at issue is systemic, not case-specific. Parliament holds the Board accountable for the CMA’s overall performance, not for the merits of individual decisions.
The Government’s argument also overlooks a key feature of the current system: Panel members are appointed by the Secretary of State.[18] The accountability chain therefore remains intact. It runs directly through the minister responsible for those appointments, who is accountable to Parliament for both the selection and oversight of panel members. The Secretary of State may remove members for incapacity, misbehaviour, or failure to perform their duties. This is the standard constitutional mechanism for ensuring democratic accountability in arm’s-length bodies, while preserving independence in individual decisions. The proposal fails to recognise that distinction.
Abolishing the independent panel without adequate safeguards raises three further concerns. First, consolidating investigative, prosecutorial, and adjudicative functions within a single institutional structure heightens the risk of confirmation bias. Second, the consultation does not address the applicable standard of appeal, despite the fact that the panel’s quasi-judicial character underpins the current deferential judicial-review standard. Third, if the Government proceeds with consolidation, it must adopt minimum structural safeguards. These should include, at a minimum, a shift to merits-based appellate review to preserve the legitimacy and predictability of the regime.
A. Concentration of Decision-Making and Confirmation Bias
The central concern is that abolishing the panel without compensating safeguards would concentrate unprecedented decision-making power in the CMA executive. The reform would strip the CMA of its defining structural feature and align it more closely with models such as the European Commission or the U.S. Federal Trade Commission. The result could be an authority that acts, in effect, as ‘judge, jury and executioner’.[19]
This consolidation heightens the risk that early analytical assumptions shape final outcomes. As commentators have observed, the proposed structure ‘risks entrenching early analytical assumptions, as a smaller group of senior officials may carry forward initial positions into final decisions’.[20] It also creates ‘a risk of CMA decision-making being more susceptible to Government pressure’ and ‘more prone to lobbying efforts and, in the long run, less predictable’.[21]
The problem is structural. When a single institution initiates an investigation, formulates a theory of harm, gathers evidence, and then adjudicates, the risk of confirmation bias increases sharply. The panel system mitigates that risk by introducing a genuinely independent, de novo layer of review. Inquiry Groups reassess both facts and theory with ‘fresh eyes’, providing a check on institutional tunnel vision.
Removing that safeguard would weaken the integrity of Phase 2 investigations. Without an independent decision-maker, Phase 2 risks becoming a procedural validation of Phase 1 conclusions, rather than a genuinely independent assessment.
B. The Standard of Appeal and Loss of Safeguards
The consultation is notably silent on the applicable standard of appeal. That omission is significant. The judicial-review standard under section 120 of the Enterprise Act rests, in part, on the quasi-judicial character of the independent panel.[22] Removing the panel removes that justification. The Government should not eliminate the structural safeguard that supports a deferential appellate standard while leaving that standard unchanged.
The Court of Appeal’s judgment in Cérélia/Jus-Rol underscores the limits of the current framework. The Competition Appeal Tribunal (CAT) must apply judicial review, even as a specialist tribunal. It may exercise a ‘high degree of scrutiny’ over CMA factual findings,[23] but it cannot substitute its own judgment on the merits. Its role remains confined to assessing rationality.
Comparative practice highlights the gap this creates. In the European Union, the General Court reviews Commission decisions for manifest errors of assessment and exercises unlimited jurisdiction over fines. The Court of Justice confirmed in CK Telecoms that the Commission must satisfy a ‘more likely than not’ standard of proof.[24] In the United States, the Department of Justice litigates before federal courts that apply de novo review. The Federal Trade Commission conducts administrative proceedings subject to ‘substantial evidence’ review. The Supreme Court’s decision in Loper Bright—overturning Chevron deference—has further strengthened judicial scrutiny of agency action.[25]
International guidance points in the same direction. The OECD’s 2021 Recommendation on Transparency and Procedural Fairness calls for access to an impartial, independent adjudicative body, with review that encompasses facts, evidence, and the grounds of decisions.[26]
A regime that concentrates investigative, prosecutorial, and adjudicative functions within a single institution, while limiting appeals to judicial review, would fall below the standard maintained in comparable jurisdictions.
C. Minimum Safeguards and the Need for Merits-Based Review
If the Government proceeds with replacing the panel system, it should adopt a minimum set of structural safeguards. These should include transparent, objective criteria for appointing members to any expert pool, combined with sufficient security of tenure to insulate them from executive pressure. Cooling-off periods should prevent recent CMA staff from joining the pool. A public conflicts register, with clear recusal triggers, should be mandatory. Crucially, a two-stage separation principle must preserve genuine independence between the Phase 1 case team and the Phase 2 decision-making body, with no overlap in personnel.
These measures address form, not substance. The more fundamental requirement is to reform the CAT’s appellate mandate. A shift from judicial review to a comprehensive merits-based standard is essential. Without it, the proposed institutional changes would lack an adequate external check.
Consolidating decision-making authority within a single body increases, rather than reduces, the need for independent review. Where internal safeguards weaken, external scrutiny must strengthen. A reform that concentrates adjudicative authority without enhancing appellate oversight remains incomplete.
III. Sunset Clauses and Time-Limited Remedies
The consultation proposes placing the CMA’s existing sunset-clause policy on a statutory footing. It would require the CMA to consider sunset clauses in all market remedies and mandate review at least once every 10 years.[27] Where the CMA declines to include a sunset clause, it must publish its reasoning.[28] This approach aligns with the Government’s pro-growth agenda. It limits legacy compliance burdens and helps ensure that remedies do not persist after market conditions change.
The economic rationale for time-limiting remedies is straightforward. Remedies impose costs: compliance costs for firms, monitoring costs for agencies, and constraints on pro-competitive, pro-consumer conduct. These costs are justified only while the underlying risk to competition persists.[29] Once that risk dissipates, the costs become deadweight losses, borne by firms and passed on to consumers through higher prices or reduced innovation. Evidence from the regulatory-burden literature suggests that compliance costs often behave as fixed overhead, rather than variable expenses. They therefore fall disproportionately on smaller firms and can create barriers to entry.[30] Sunset clauses address this problem by reversing the burden of justification. Rather than requiring firms to petition for removal, they require the authority to demonstrate that continued intervention remains necessary.
International practice supports this approach. In the United States, the Federal Trade Commission adopted a formal sunset policy in 1995, establishing automatic 20-year termination for administrative orders.[31] Department of Justice merger consent decrees have incorporated 10-year sunset provisions as standard practice since 1979, and recent FTC merger orders follow similar timelines.[32] In the European Union, behavioural commitments under Article 9 of Regulation 1/2003 are typically limited to five to 10 years.[33] The OECD and the ICN Merger Remedies Guide likewise endorse time-limiting as standard practice.[34]
Domestic experience points in the same direction. The CMA’s strategic review of market remedies reflects a growing recognition that existing measures often outlast their usefulness.[35] The Restriction on Agreements and Conduct (Tour Operators) Order 1987 and the Foreign Package Holidays (Tour Operators and Travel Agents) Order 2001 illustrate the problem. Both targeted a market dominated by high-street travel agents and imposed obligations relating to in-store price display, resale price maintenance, and tied sales of travel insurance.[36] The shift to online booking rendered many of these requirements obsolete. Firms nevertheless continued to incur compliance costs for decades, including nearly 40 years under the 1987 Order.
The Payment Protection Insurance (PPI) Market Investigation Order 2011 provides a further example.[37] The Order imposed a detailed compliance regime, including bans on point-of-sale PPI, mandatory annual customer reminders, claims-ratio disclosures, compliance officer requirements, and monitoring obligations such as mystery shopping and reporting. These measures responded to a market characterised by widespread mis-selling and substantial consumer harm. That market has since disappeared. The Financial Conduct Authority set an August 2019 deadline for PPI complaints, by which point more than £38 billion in redress had been paid.[38] The FCA has confirmed that PPI is no longer sold alongside credit products and has proposed removing PPI-specific rules on the basis that they no longer serve a regulatory purpose.[39] Despite this, the CMA’s 2011 Order remains formally in force. Its compliance framework continues to apply, at least nominally, to a market that has materially evolved. This illustrates how remedies can persist in the absence of mechanisms that compel periodic reassessment.
The proposal would address part of this problem, but further refinements would improve its effectiveness. First, a risk-based review cadence would be preferable to a uniform 10-year default. In fast-moving digital and technology markets, competitive conditions can change far more quickly. A shorter review cycle—such as five years, consistent with the UK’s SMS regime—would better reflect those dynamics. Differentiated defaults, with shorter periods for fast-cycle markets and longer ones for more stable sectors, would improve calibration.
Second, the Government should establish clear petition mechanisms that allow affected parties to request review. These mechanisms should include defined evidentiary standards and binding response timelines. Third, the CMA’s existing transparency practices should be codified. The authority should publish its reasoning in all cases, whether it removes, maintains, or modifies a remedy. This would strengthen accountability and allow firms to understand the analytical basis for continued intervention.
IV. Concurrency and the Risk of Fragmentation
The consultation seeks views on whether sector regulators should oversee remedies imposed by the Competition and Markets Authority (CMA), whether the proposed consultative approach to concurrency is appropriate, and whether further reform is needed.[40]
If not carefully designed, expanded concurrency risks multiplying intervention without improving outcomes. It may undermine the Government’s objectives of pace, predictability, proportionality, and process. The UK’s existing framework already grants competition-law powers to sector regulators, including Ofcom, Ofgem, and the Financial Conduct Authority.[41] While this model can draw on sector-specific expertise, it also creates well-documented risks of duplication, forum shopping, and inconsistent decision-making.
These risks would extend—and intensify—if sector regulators were permitted to oversee CMA-imposed remedies. Remedies are not merely technical follow-ons to liability findings. They are ongoing, market-shaping interventions. Splitting responsibility for their design, implementation, and review across multiple authorities risks inconsistent interpretation, higher compliance costs, and diffused accountability. From a law & economics perspective, such fragmentation increases error costs. It raises the likelihood of overbroad or misapplied remedies and makes those errors harder to detect and correct.
The OECD has warned that concurrency frameworks can generate overlap and inefficiency where roles are not clearly defined.[42] The Strategic Steer reinforces this point. It directs the CMA to prioritise interventions with a ‘clear and direct UK impact’ and to consider whether parallel enforcement elsewhere can address concerns.[43] This principle should apply with equal force within the United Kingdom’s own institutional framework. That principle should apply equally within the UK. A system that allows multiple domestic authorities to exercise overlapping powers risks replicating, at home, the duplication the Government seeks to avoid abroad.
Reform should therefore follow an explicit anti-duplication principle. Where the CMA imposes remedies, it should remain the lead authority responsible for oversight and review. Sector regulators can and should contribute expertise, but they should not exercise parallel supervisory authority absent a clear statutory justification grounded in necessity and proportionality. This approach would promote predictability by ensuring that firms face a single decision-maker and a single set of binding obligations.
A consultative model offers the right balance. Sector regulators possess valuable technical expertise, particularly in telecommunications, energy, and financial services. They should provide evidence, market knowledge, and technical input. That does not require shared authority over remedies. Granting such authority would risk turning competition enforcement into co-regulation, blurring the line between competition law and sector-specific regulation. This risk is especially acute given the consultation’s broader trajectory, which already expands the CMA’s quasi-regulatory role.
A well-designed concurrency regime must also preserve clear lines of accountability. It should specify which authority designs remedies, which monitors compliance, and which is accountable for outcomes, including effects on investment, innovation, and consumer welfare. Any reform should therefore require designation of a single lead authority, transparent allocation of responsibilities, and unified review mechanisms. These should include sunset clauses and periodic reassessment conducted by, or under the authority of, the lead institution.
As the CMA’s powers expand, the need for clarity and accountability increases. Concurrency reform should reduce fragmentation, not entrench it.
V. Expansion of Algorithmic Investigative Powers
Chapter 4 of the consultation proposes extending the DMCC Act’s investigative powers beyond the Strategic Market Status regime to general competition enforcement.[44] These powers include the ability to require firms to explain how algorithms operate, conduct tests, generate data, and perform specified demonstrations. In its March 2026 response, the CMA went further, arguing that such powers ‘should also not be limited to algorithms’ but should apply to all investigative information-gathering across the economy.[45]
This proposal is both disproportionate and without clear precedent. It also risks undermining the predictability that the Government has placed at the centre of its reform agenda.
No comparable jurisdiction has taken this approach. In the European Union, algorithm-specific obligations remain confined to the Digital Markets Act, which applies only to designated gatekeepers, and the AI Act, which targets high-risk systems. Competition enforcement under Articles 101 and 102 TFEU continues to rely on established investigative tools.[46] In the United States, the Federal Trade Commission uses existing powers under Sections 5 and 9 to investigate algorithmic conduct, while the Department of Justice pursues such cases under the Sherman Act without additional statutory authority.
The UK proposal would go significantly further. It would extend algorithmic-testing powers from a targeted ex ante digital-markets regime to general competition enforcement across the entire economy. No other major jurisdiction has adopted, or proposed, such an expansive approach.
A. Sector-Specific Powers in a General Enforcement Context
Section 69 of the DMCC Act was designed for a narrow, ex ante regime applied to a small number of firms designated with Strategic Market Status. Extending these powers to general competition enforcement would subject firms across the entire economy to quasi-regulatory oversight developed for a fundamentally different context.
The distinction is critical. Existing regimes that mandate algorithmic testing—such as financial trading under the Financial Services Act, medical devices under MHRA guidance, and autonomous vehicles under the Automated Vehicles Act 2024—target specific, well-defined harms, including systemic financial risk, patient safety, and road safety.[47] They operate within cooperative regulatory frameworks, where firms accept oversight as a condition of market access. Regulators in these sectors possess deep, domain-specific expertise, and the scope of regulated algorithms is clearly defined.
General competition enforcement differs in each respect. It is inherently adversarial. The harms it addresses are diffuse and often contested. There is no clear limiting principle to define which algorithms fall within scope; in practice, any automated decision-making system in any sector could be captured. The CMA would therefore need to assess algorithmic systems across the entire economy—from financial services and logistics to healthcare and retail—an institutional demand that no competition authority is equipped to meet.
The burden would not fall evenly. As Mayer Brown observed in 2021, even ‘smaller firms licensing “off the shelf” algorithms would face the same investigatory expectations as larger/dominant firms’ and may lack access to the underlying data needed to comply. Extending these powers risks imposing disproportionate obligations on firms least able to meet them, without a corresponding gain in enforcement effectiveness.[48]
B. Trade Secrets and Confidentiality Risks
Algorithms, training weights, and machine-learning architectures are among firms’ most valuable trade secrets. As Katarina Foss-Solbrekk observes, trade secrets are ‘the only option left for firms wishing to protect their algorithms’[49]—a trend reinforced by Robin Feldman’s work showing that post-Alice Corp. patent law has pushed developers further toward secrecy.[50] Compelling disclosure of such assets in an adversarial investigation creates significant security risks. In merger reviews involving AI firms, the CMA could gain access to some of the most commercially sensitive intellectual property in the world.
These risks are not hypothetical. During the Microsoft/Activision Blizzard merger review, unredacted confidential materials—including Xbox console roadmaps, next-generation hardware plans, unannounced game titles, and internal acquisition-target lists—were inadvertently disclosed on a public court server in related FTC proceedings.[51] The episode illustrates how even well-resourced authorities can fail to fully protect sensitive information.
Recent UK litigation reinforces the point. In Tereos v Competition & Mkts. Auth., proceedings before the Competition Appeal Tribunal highlighted the difficulty of balancing confidentiality interests under the current framework.[52] Expanding algorithmic testing powers would intensify these challenges, increasing both the volume and sensitivity of information subject to disclosure.
C. Institutional Limits and Practical Risks
Competition authorities are not institutionally equipped to oversee algorithmic systems in a quasi-regulatory capacity. Algorithms are dynamic and continuously adaptive, not static rule-based systems. Requiring firms to run ‘specified tests’ under artificial conditions risks producing outputs that bear little resemblance to real-world market behaviour.
The Hayekian knowledge problem is directly relevant. Central authorities lack the localised, context-specific information needed to assess whether algorithmic outcomes are ‘fair’ or ‘anticompetitive’.[53] The OECD’s 2023 Algorithmic Competition Roundtable found that responses ‘generally suggested that most algorithmic theories of harm are already captured by existing law’.[54] Former US FTC Chair Maureen Ohlhausen has likewise argued that ‘concerns about algorithms are a bit alarmist’ and that ‘expanding use of algorithms raises familiar issues well within the existing canon’.[55] These assessments weigh against introducing novel, intrusive investigative powers where existing tools already suffice.
The practical consequences for merger control are significant. Merger review operates under strict statutory deadlines. Adding complex algorithmic testing and demonstration requirements would slow proceedings, increase costs for merging parties, and introduce further uncertainty into already demanding timelines. At the margin, such burdens may deter investment in the UK, contrary to the Government’s stated objectives.
The CMA’s own 2026–2029 strategy recognises this risk, warning that ‘burdensome interventions risk stifling innovation’.[56] Extending algorithmic investigation powers across general competition enforcement would exemplify precisely that type of intervention.
VI. Conclusion
ICLE welcomes the Government’s commitment to refining the United Kingdom’s competition regime. Several proposals in the consultation—notably the statutory presumption of sunset clauses and the extension of Phase 1 remedy timeframes—reflect sound, evidence-based policymaking. They will reduce regulatory friction and provide the legal certainty required for sustained economic growth.
Several of the most consequential proposals, however, risk undermining those objectives. Abolishing the CMA’s independent panel without introducing merits-based appellate review at the Competition Appeal Tribunal would concentrate investigative, prosecutorial, and adjudicative authority within a single institution. That shift would erode structural safeguards that underpin the credibility of UK competition enforcement and place the regime below the standard maintained in comparable jurisdictions.
Extending algorithmic investigation powers from the DMCC Act’s tightly circumscribed Strategic Market Status regime to general competition enforcement would be internationally unprecedented and disproportionate to any demonstrated need. It also risks chilling the algorithm-driven innovation on which the Government’s growth agenda depends. Similarly, permitting sector regulators to exercise parallel supervisory authority over CMA-imposed remedies would fragment accountability and increase compliance burdens. The CMA should remain the single lead authority, with sector regulators contributing expertise through a consultative model.
A recurring concern throughout this submission is that many proposals draw on the DMCC Act’s digital-markets regime as an institutional template. That framework received Royal Assent in May 2024, took effect in January 2025, and has so far been applied to only three Strategic Market Status designations across two firms. Extending its most intrusive features before they have been tested in practice risks compounding the proportionality and predictability concerns reflected in both the Strategic Steer and the CMA’s ‘4Ps’ framework.
Reform should also be guided by the error-cost framework. False positives in competition enforcement are typically more costly and less self-correcting than false negatives, particularly in innovation-intensive markets. Expanding CMA powers therefore requires stronger evidentiary standards, more robust appellate review, and mandatory sunset provisions—not lower thresholds for intervention.
The Government’s ambition to create a ‘best-in-class’ competition regime calls for a selective, not sweeping, reform agenda. Targeted procedural improvements should be prioritised. Expanding investigatory reach in ways that increase uncertainty, deter investment, and weaken predictability would ultimately undermine the pro-growth mission at the centre of the Government’s programme.
[1] Dep’t for Bus. & Trade, Refining Our Competition Regime: Driving Growth and Enhancing Competition for Businesses and Consumers (20 Jan. 2026) [hereinafter Consultation].
[2] Sarah Cardell, New CMA Proposals to Drive Growth, Investment and Business Confidence, Competition & Mkts. Auth. Blog (13 Feb. 2025), https://competitionandmarkets.blog.gov.uk/2025/02/13/new-cma-proposals-to-drive-growth-investment-and-business-confidence.
[3] Dep’t for Bus. & Trade & Competition & Mkts. Auth., Strategic Steer to the Competition and Markets Authority (15 May 2025), https://www.gov.uk/government/publications/strategic-steer-to-the-competition-and-markets-authority/strategic-steer-to-the-competition-and-markets-authority [hereinafter Strategic Steer].
[4] HM Gov’t, Kickstarting Economic Growth (last visited 26 Mar. 2026), https://www.gov.uk/missions/economic-growth.
[5] Keir Starmer, Keir Starmer: Growth Is the Chancellor’s Priority, The Times (28 Jan. 2025), https://www.thetimes.com/uk/politics/article/keir-starmer-growth-chancellor-k68ptvh6x; HM Treasury & Rachel Reeves, Chancellor Calls on Watchdog Bosses to Tear Down Regulatory Barriers That Hold Back Growth (22 Jan. 2025), https://www.gov.uk/government/news/chancellor-calls-on-watchdog-bosses-to-tear-down-regulatory-barriers-that-hold-back-growth.
[6] Guy Taylor, Ousted CMA Chair Bokkerink Defends Tenure, City A.M. (8 Feb. 2025), https://www.cityam.com/ousted-cma-chair-bokkerink-defends-tenure.
[7] Dep’t for Bus. & Trade, Competition & Mkts. Auth. & Jonathan Reynolds, Former Amazon Boss Named Interim Chair of CMA (21 Jan. 2025), https://www.gov.uk/government/news/former-amazon-boss-named-interim-chair-of-cma.
[8] See Strategic Steer, supra note 3.
[9] Cardell, supra note 2.
[10] Competition & Mkts. Auth., Mergers Charter: How to Work with the CMA on a Merger Investigation (12 Mar. 2025), https://www.gov.uk/government/publications/mergers-charter-how-to-work-with-the-cma-on-a-merger-investigation/mergers-charter.
[11] Dep’t for Bus. & Trade, Regulator Dashboard (21 Oct. 2025), https://www.gov.uk/government/publications/regulator-dashboard.
[12] Consultation, supra note 1, at 3 (‘improving the pace, predictability, proportionality, and process of engagement through refinements to the legislative framework for competition’).
[13] Richard Feasey, Address at the Law Council of Australia’s Competition and Consumer (ACCC) Annual Conference: The Role of the CMA Panel in Decision Making: Merger Enforcement and Reform (1 Sept. 2021), https://www.gov.uk/government/speeches/the-role-of-the-cma-panel-in-decision-making-merger-enforcement-and-reform.
[14] Consultation, supra note 1, ¶ 25.
[15] Id.
[16] Id. ¶ 6.
[17] See Gavin Robert, Three Years of CMA Merger Control: A Statistical Review, 16 Competition L.J. 211, 216 (2017) (‘…until the CMA formed as a single authority in 2014, the UK maintained separate Phase 1 and Phase 2 bodies: the OFT decided Phase 1, and, on reference, the CC conducted Phase 2. Many viewed this as an effective—if costly—check on confirmation bias.’).
[18] Consultation, supra note 1, ¶ 26.
[19] UK Government Consults on Major Reforms to Competition Processes, Osborne Clarke (4 Feb. 2026), https://www.osborneclarke.com/insights/uk-government-consults-major-reforms-competition-processes.
[20] UK Government Launches Consultation Overhauling the CMA’s Competition Regime, Baker Botts (3 Feb. 2026), https://www.bakerbotts.com/thought-leadership/publications/2026/february/uk-government-launches-consultation-overhauling-the-cmas-competition-regime.
[21] Greg Dowell & Fiona Beattie, Refining the UK Competition Regime: Proposals to Change Mergers and Markets Investigations, Macfarlanes (30 Jan. 2026), https://www.macfarlanes.com/insights/102mfb9/refining-the-uk-competition-regime-proposals-to-change-mergers-and-markets-inves.
[22] Cérélia Group Holdings SAS v Competition & Mkts. Auth. [2024] EWCA Civ 352, paras 28–41.
[23] Id.
[24] Case C-376/20 P, CK Telecoms UK Invs. Ltd v Comm’n, ECLI:EU:C:2023:561 (2023).
[25] Loper Bright Enters. v Raimondo, 603 U.S. 369 (2024).
[26] Org. for Econ. Co-operation & Dev. (OECD), Recommendation of the Council on Transparency and Procedural Fairness in Competition Law Enforcement (2021), https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0465.
[27] Consultation, supra note 1, ¶ 26.
[28] Id.
[29] Org. for Econ. Co-operation & Dev. (OECD), Remedies and Commitments in Abuse Cases, OECD Roundtables on Competition Policy Papers No. 288, at 14 (Nov. 2022), https://www.oecd.org/content/dam/oecd/en/publications/reports/2022/11/remedies-and-commitments-in-abuse-cases_c215224a/b975b0e3-en.pdf (‘An early termination ensures the remedy remains proportionate—it ceases once unnecessary—and frees the authority from ongoing monitoring, so it can redeploy resources more effectively.’).
[30] Joseph J. Cordes, Susan E. Dudley & Layvon Q. Washington, Regulatory Compliance Burdens: Literature Review & Synthesis, Geo. Wash. Univ. Regulatory Stud. Ctr., at 14–16 (Oct. 2022), https://regulatorystudies.columbian.gwu.edu/sites/g/files/zaxdzs4751/files/2022-10/regulatory_compliance_burdens_litreview_synthesis_finalweb.pdf.
[31] U.S. Fed. Trade Comm’n, Policy Statement on the Duration of Competition Orders (1995), https://www.ftc.gov/system/files/documents/public_statements/410481/frndurationoforders.pdf (establishing an automatic 20-year sunset for administrative orders).
[32] U.S. Dep’t of Justice, Antitrust Div., Merger Remedies Manual (2020), https://www.justice.gov/atr/page/file/1312416/dl; see also In re Synopsys, Inc., FTC Dkt. No. C-4820, at 19 (2025), https://www.ftc.gov/legal-library/browse/cases-proceedings/241-0059-c-4820-synopsys-inc-ansys-inc-matter (imposing a ten-year consent order).
[33] EU behavioural commitments under Article 9 of Regulation 1/2003 are typically time-limited, usually five to ten years depending on the concern. See, e.g., Commission Decision of 16 Dec. 2009, Case COMP/C-3/39.530 — Microsoft (Tying) (five-year commitment); Commission Decision of 29 Apr. 2014, Case AT.39939 — Samsung — Enforcement of UMTS Standard Essential Patents (five-year commitment); Commission Decision of 11 July 2024, Case AT.40452 — Apple — Mobile Payments, C(2024) 4761 (ten-year commitment).
[34] Org. for Econ. Co-operation & Dev. (OECD), Remedies and Commitments in Abuse Cases, supra note 29; Int’l Competition Network (ICN), Merger Working Grp., Merger Remedies Guide, at 4 (2016), https://www.internationalcompetitionnetwork.org/wp-content/uploads/2018/05/MWG_RemediesGuide.pdf.
[35] Competition & Mkts. Auth., Strategic Review of CMA Markets Remedies (19 Jan. 2026), https://www.gov.uk/cma-cases/strategic-review-of-cma-markets-remedies; Consultation, supra note 1, at 3 (‘For each remedy, the CMA expects that clear, straightforward changes in circumstances may render the measure unnecessary, allowing it to be amended or withdrawn.’).
[36] The Restriction on Agreements and Conduct (Tour Operators) Order 1987, SI 1987/1131, https://www.legislation.gov.uk/uksi/1987/1131/made; The Foreign Package Holidays (Tour Operators and Travel Agents) Order 2001, SI 2001/2581, https://www.legislation.gov.uk/uksi/2001/2581/contents/made.
[37] Competition & Mkts. Auth., PPI Market Investigation Order 2011 (24 Mar. 2011), https://assets.publishing.service.gov.uk/media/5c94dcfce5274a5e10fd5944/PPI_Order.pdf.
[38] Fin. Conduct Auth., PPI Complaints (updated 23 Apr. 2025), https://www.fca.org.uk/consumers/ppi-complaints.
[39] Fin. Conduct Auth., FCA Strips Back Insurance Rulebook (14 May 2025), https://www.fca.org.uk/news/press-releases/fca-strips-back-insurance-rulebook.
[40] Consultation, supra note 1, questions 11–13.
[41] Enterprise and Regulatory Reform Act 2013, § 51 & sch. 14; Competition Act 1998, § 54.
[42] Org. for Econ. Co-operation & Dev. (OECD), Competition Law and Policy in the United Kingdom (Peer Review, 2020).
[43] HM Gov’t, Strategic Steer to the Competition and Markets Authority, supra note 3.
[44] Digital Markets, Competition and Consumers Act 2024, c. 13, § 69 (UK).
[45] Competition & Mkts. Auth., Response to the Consultation on Refining Our Competition Regime (Mar. 2026) (‘algorithmic investigation powers should not be limited to algorithms’).
[46] See Regulation (EU) 2022/1925 (Digital Markets Act); Regulation (EU) 2024/1689 (AI Act) (neither extends algorithmic investigation powers to general competition enforcement under Articles 101–102 TFEU).
[47] Directive 2014/65/EU, art. 17 (MiFID II) (algorithmic trading obligations); U.S. Food & Drug Admin. & MHRA, Good Machine Learning Practice for Medical Device Development (2021); Automated Vehicles Act 2024 (UK).
[48] Mayer Brown LLP, Response to CMA Consultation on Algorithms (2021).
[49] Katarina Foss-Solbrekk, Three Routes to Protecting AI Systems and Their Algorithms Under IP Law: The Good, the Bad, and the Ugly, 16 J. Intell. Prop. L. & Prac. 1300 (2021).
[50] Robin Feldman, Patent Demands & Startup Companies: The View from the Venture Capital Community, 16 Nev. L.J. 583 (2016).
[51] Cecilia D’Anastasio & Leah Nylen, Microsoft Mistakenly Reveals Secret Game Plans in Court Case, Bloomberg (19 Sept. 2023).
[52] Tereos v. CMA [2024] CAT (UK) (demonstrating difficulty of balancing confidentiality interests in CMA proceedings).
[53] See Friedrich A. Hayek, The Use of Knowledge in Society, 35 Am. Econ. Rev. 519, 524 (1945).
[54] Org. for Econ. Co-operation & Dev. (OECD), Algorithmic Competition (Roundtable Background Paper, 2023) (‘responses generally suggested that most algorithmic theories of harm are already captured by existing law’).
[55] Maureen K. Ohlhausen, Should We Fear the Things That Go Beep in the Night? Some Initial Thoughts on the Intersection of Antitrust Law and Algorithmic Pricing, Remarks at the Concurrences Antitrust in the Financial Sector Conference (23 May 2017).
[56] Competition & Mkts. Auth., Annual Plan 2026–29 (2026) (‘[b]urdensome interventions risk stifling innovation’).