Antitrust and Competition Technology Year in Review 2025

As 2025 progressed, competition law enforcement became increasingly policy-driven, influenced by domestic political agendas and transatlantic tensions with direct consequences for dealmaking, compliance, and regulatory risk. 

In the EU, the European Commission’s review of its Horizontal and Non-Horizontal Merger Guidelines signalled a push to modernise merger control for today’s economic realities, while enforcement of the Digital Markets Act (DMA) moved into a more assertive phase. In the UK, 2025 was defined by a recalibration of enforcement priorities, combining expanded powers under the Digital Markets, Competition and Consumers Act (DMCC Act) with political pressure, reflected in the government’s May 2025 steer and a change in CMA leadership to align competition policy more closely with growth and competitiveness.

US antitrust enforcement in 2025 paired continued pressure on big tech conduct with a more pragmatic stance on transactions, particularly in AI and other innovation driven markets. The DOJ and FTC continued to pursue high-profile litigation, but senior officials increasingly framed antitrust as traditional law enforcement grounded in traditional theories of harm rather than an open-ended regulatory program, echoing the position in the UK. That shift was reflected in the absence of challenges to prominent tech, AI, and vertical deals. 

For the technology sector, 2025 foreshadows the landscape in 2026, with sustained scrutiny of platform power, accelerating digital enforcement, and an expanding regulatory perimeter reshaping both regulatory engagement and transactional risk.

Merger Control in Transition

UK Merger Control: A Year of Recalibration

UK merger control shifted markedly in 2025. Following the sudden replacement of CMA Chair Marcus Bokkerink with Doug Gurr in January, and against the backdrop of the government’s May 2025 policy steer, the CMA signalled an intent to intervene less on global mergers and adopt a more UK-focused, pro-investment approach.

The change is evident in outcomes. In 2025, the CMA blocked no mergers — the first such year since 2017 — and the number of cases cleared with remedies fell to six, from seven in 2024 and 12 in 2023. It may be too soon to label this a firm trend; outcomes depend on the deal mix and do not capture transactions abandoned by merging parties preemptively. But direction of travel seems clear: The government’s pro-growth agenda has materially shaped UK merger enforcement in 2025.

Substantively, the CMA embedded its ‘4Ps’ framework — pace, predictability, proportionality, and process — across merger control. The revised Merger Remedies Guidance, published in December, signals a more pragmatic approach, recognising that well-designed behavioural remedies may, in limited cases, provide a credible alternative to structural solutions, including at Phase 1. Structural remedies nonetheless remain the default, as illustrated by the rejection of Phase 1 behavioural commitments in Getty Images/Shutterstock.

Procedural reform accompanied this shift. Updated jurisdiction and procedure guidance introduced a 40-working-day prenotification KPI, more structured engagement with parties, and clearer limits on the share-of-supply test by anchoring it to statutory criteria and competition-relevant goods or services. Further reform will follow in 2026, with the government announcing a consultation earlier this month into potential changes to Phase 2 decision-making that would concentrate decision-making authority at the CMA’s board level — heightening the importance of early engagement and a disciplined UK-specific narrative in complex cases.

EU Merger Guidelines Under Review

The CMA was not alone in recalibrating merger control. In parallel, the European Commission launched a review of its Horizontal and Non-Horizontal Merger Guidelines, last updated in 2004 and 2007 and increasingly misaligned with today’s digital, geopolitical, and industrial policy realities.

Following a public consultation from May to September 2025, the review — echoing the Draghi Report — aims to make EU merger control ‘fit for purpose.’ Some proposals codify analytical approaches already evident in recent cases or align with broader enforcement trends, including a renewed emphasis on growth and innovation-led productivity; others mark a more substantive shift.

Notably, the Commission suggests that environmental, scale-up, or security-of-supply efficiencies may, in certain cases, offset competitive harm, signalling a modest softening of its traditionally narrow approach to efficiency defences. This reflects a post-Draghi view that merger control can facilitate pro-competitive transactions advancing sustainability or competitiveness objectives, particularly in strategic sectors. The Commission remains cautious, however, reiterating that acquisitions of nascent competitors by large incumbents will continue to attract heightened scrutiny.

The review also proposes rebuttable presumptions to identify potentially problematic mergers alongside existing safe harbours. While this may enhance predictability, it could shift the burden onto parties to demonstrate the absence of anticompetitive effects. With the consultation have closed in September 2025 and the substantive EUMR test unchanged, the practical impact will only become clear once the revised guidelines are published in 2026.

For the technology sector — especially AI, cloud, and compute-intensive services — the review signals a broader reframing of competition policy around scale, innovation, and strategic autonomy. Access to capital and complementary assets is increasingly recognised not only as a competition risk but also as a potential enabler of innovation and resilience, provided foreclosure concerns can be effectively addressed.

US Merger Enforcement as Law Enforcement, Not Regulation

US leadership messaging in 2025 consistently emphasised a return to traditional law-enforcement principles in antitrust enforcement. In her first major address, Assistant Attorney General Gail Slater stressed that US antitrust is ‘law enforcement, not regulation’, cautioning against remedies that substitute ongoing oversight for proving concrete competitive harm. Deputy Assistant Attorney General Bill Rinner echoed the point in June, underscoring that the Division’s civil merger program must respect the Division’s role as a ‘law enforcer protecting free markets, not as a regulator’ and warning against treating ‘absolute deterrence’ as a guiding aim of merger policy. 

The point about deterrence is notable for what it implicitly rejects. It contrasts sharply with comments made by the prior FTC leadership under Lina Khan, which had embraced deterrence as a metric of success, celebrating deals ‘no longer making it out of the boardroom’. The change in messaging illustrates a meaningful tonal shift: continued vigilance where facts support intervention but a clearer effort to avoid chilling dealmaking as a policy objective.

This backdrop helps explain the government’s posture toward AI transactions in 2025. The FTC’s most detailed assessment of AI partnerships this year came from the outgoing administration: a January 2025 staff report framing AI partnerships as potential competitive flashpoints, highlighting theories that such arrangements could create dependency and lock-in, raise switching costs, restrict access to key inputs, and enable access to competitively sensitive information — issues the prior leadership viewed as grounds for more interventionist vertical enforcement.

But the accompanying dissents by Commissioners Ferguson and Holyoak foreshadowed a change in approach. Both endorsed scrutiny of incumbents using partnerships to blunt emerging threats yet warned against ‘charging headlong to regulate AI’, stressing the risk of stifling innovation or driving development abroad. Holyoak similarly highlighted the report’s limits and objected to portions she viewed as policy signalling beyond the factual record — an implicit push away from the prior administration’s more sceptical framing of AI relationships as presumptively suspect.

Taken together, the report placed AI partnerships firmly on the enforcement agenda, while the dissents pointed toward a ‘first, do no harm’ approach — continued scrutiny but grounded in clearer theories of harm and a reluctance to use antitrust as a proxy for AI regulation.

AI Acquisitions and Vertical-Leaning Deals — Intensive Interest but Limited US Enforcement

The largest technology transactions over the past year were notable in the US less for enforcement outcomes but rather for what their trajectories suggest about the antitrust agencies’ priorities. Several headline deals drew meaningful attention and, in some cases, reportedly underwent in-depth reviews. Yet, in the United States, that scrutiny largely failed to translate into challenges. Instead, the market saw a pattern aligned with the aforementioned ‘do no harm’ enforcement posture.

This dynamic was visible across the year’s most prominent transactions. Google’s proposed acquisition of Wiz — widely viewed as a litmus test of how the US agencies would approach a major platform acquiring a fast-growing, independent security provider — was reportedly reviewed by the DOJ but ultimately proceeded without a US enforcement action. Salesforce’s acquisition of Informatica similarly represented a potential case of big tech entrenchment in AI, but likewise, it did not draw a challenge in the United States. Meta’s investment in Scale AI, paired with its high-profile leadership transition, illustrates the way AI dealmaking increasingly blends investment, commercial alignment, and talent acquisition in ways that raise competition questions without fitting neatly into a traditional acquisition framework.

What connects these transactions is not size but structure. Each is, substantially, a vertical combination aimed at securing inputs, infrastructure, distribution, or data-layer advantages that are strategically important in AI. 

Under the previous administration’s more sceptical view of vertical integration and platform ecosystems, these deals might have been positioned as marquee enforcement opportunities. This year, however, even where deals were subject to substantial reviews, outcomes in the United States were comparatively restrained: Scrutiny did not become litigation. The overall pattern reinforces the sense that agencies are reluctant to use AI era dealmaking as a vehicle for aggressive, market shaping enforcement unless supported by a traditional theory of harm likely to succeed in court.

HPE/Juniper: Novel Remedies and a Contentious Tunney Act Review

HPE/Juniper began as a conventional horizontal challenge but evolved into something far more complex: a novel remedy package followed by unusually contentious Tunney Act proceedings. 

On 30 January 2025, the DOJ sued to block HPE’s proposed acquisition of Juniper, alleging harm to competition in the US market for enterprise-grade wireless LAN (WLAN) solutions. Less than two weeks before the trial, the DOJ announced a settlement requiring HPE to divest its Instant On business and adopt licensing commitments relating to Juniper’s software.

What made the matter stand out was not the remedy itself but the public controversy surrounding the settlement process. Press reports described allegations that HPE and Juniper (or their advocates) sought a ‘sweetheart’ resolution by lobbying senior DOJ officials outside the Antitrust Division chain of command, including officials above Assistant Attorney General Gail Slater. Those allegations became a rallying point for critics who argued that the relief DOJ accepted did not match the competitive harms alleged in the complaint and that the settlement process lacked transparency.

Those concerns have now surfaced in Tunney Act litigation. In October 2025, a coalition of state attorneys general and the District of Columbia moved to intervene in the Tunney Act review, specifically alleging that the settlement was the product of undue influence by well connected lobbyists and urging the court to scrutinise both the adequacy of the remedy and the circumstances in which it was negotiated. In November 2025, Judge P. Casey Pitts granted the states’ motion to intervene over the objections of the DOJ and the parties.

For dealmakers, HPE/Juniper is a case study on two levels. Substantively, it suggests that creative remedies, including combining divestitures with forward-looking behavioural commitments, are a potential path to resolution in the face of traditional horizontal competitive concerns. Procedurally, it underscores that the risk for merger parties may not end with a DOJ settlement: In high profile matters, Tunney Act review can become a second stage of litigation, particularly where state enforcers diverge from federal agencies in their assessment of the deal.

European Authorities Heighten Scrutiny of ‘Below-Threshold’ Deals

A defining theme carried over from prior years was continued uncertainty over the reliability of traditional notification thresholds in assessing filing obligations and antitrust risk. Although the European Court of Justice’s decision in Illumina/Grail curtailed the Commission’s ability to review below-threshold transactions under Article 22 EUMR, the Commission’s renewed push for Member States to adopt national call-in powers signals that such deals remain firmly on its radar.

The Commission’s position is that transactions falling below EU thresholds may still be referred under Article 22 where they fall within the scope of a Member State’s discretionary call-in regime — an interpretation not yet endorsed by the EU courts. That issue came into sharp focus in early 2025 when NVIDIA challenged the Commission’s acceptance of an Article 22 referral in NVIDIA/RUN:AI, arguing that reliance on Italy’s ex-post, discretionary call-in powers infringes fundamental EU law principles, including legal certainty, proportionality, and equal treatment, and is incompatible with Illumina/Grail.

This area is likely to remain in flux through 2026. While the forthcoming judgment in NVIDIA/RUN:AI should clarify the limits of the Commission’s Article 22 powers, increased national enforcement is also likely as Member States expand and test their call-in regimes. For dealmakers, the message is clear: Transactions involving strategically significant or national security–sensitive assets, serial acquisitions, or potential ‘killer acquisitions’ will continue to attract scrutiny. Early regulatory risk assessment — and careful alignment of timelines and risk allocation in transaction documents — will be critical to avoiding unwelcome surprises.

Major Developments in the Digital Markets

Google Search Remedies: A Behavioural Decree Focused on Distribution, Data, and ‘Next-Gen’ Defaults

The remedies phase of the Google Search monopolisation case underscored a theme likely to recur in litigation against major technology platforms: The hard work begins after determining liability, when the court must craft forward-looking, administrable relief in a rapidly evolving market. In a detailed remedies opinion issued in September, Judge Mehta rejected the plaintiffs’ headline structural proposals — most notably a Chrome divestiture and a contingent Android divestiture — instead imposing a sweeping package of behavioural restrictions aimed at the distribution mechanisms the court found had entrenched Google’s search monopoly.

At the centre of the order is a comprehensive set of rules designed to unwind exclusivity and reduce contractual lock in at key search access points. The court barred Google from entering or maintaining exclusive contracts relating to the distribution of Google Search, Chrome, Google Assistant, and the Gemini app and prohibited Google from conditioning the licensing of the Google Play Store (or other Google apps) on the placement or preloading of those products. The order further restricts Google from using revenue share or other consideration to tie the placement of one Google product to another and prevents Google from conditioning payments on maintaining default status or placement for more than one year — effectively forcing more frequent renegotiation in the distribution channels the court deemed competitively critical in its liability opinion. Consistent with that targeted approach, the court declined to ban distribution payments outright, reasoning that a blanket prohibition could cause downstream harms to partners, related markets, and consumers.

Beyond these contracting constraints, the order imposes ‘access’ remedies intended to lower barriers to entry and help rivals scale. Google must make certain search index and user interaction data available to ‘Qualified Competitors,’ with the scope of sharing narrowed to align with the conduct the court found unlawful and to mitigate privacy and implementation concerns. Google is also required to offer search and search text ads syndication services, enabling rivals to deliver competitive quality results and ads while building their own search technologies and capacity. To support implementation, the order sets out a multiyear compliance structure, including the establishment of a technical committee and related oversight processes, with the court concluding that a six year term is appropriate to operationalise the data sharing, syndication, and monitoring regime.

Ultimately, the remedies decision landed where the liability theory began: Exclusive distribution arrangements were the backbone of the government’s case, and unwinding those arrangements became the core of the relief. The strategic takeaway for future tech sector cases is significant. Even after a liability win, courts tend to favour remedies narrowly tailored to the specific exclusionary conduct proved, rather than broad structural reconfigurations of a defendant’s business.

Google has appealed the underlying monopoly ruling and has sought to pause implementation of key remedies while the appeal proceeds.

Google Ad Tech Liability: A Second Major US Government Win

In April, the DOJ and several states prevailed on liability in their ad technology monopolisation case against Google, with Judge Brinkema concluding that Google unlawfully maintained monopoly power in two publisher-side markets: open-web display publisher ad servers and open-web display ad exchanges. The court did not accept the US government’s attempt to prove a separate advertiser ad network monopolisation theory, narrowing the case to the publisher-facing layers of the stack.

The liability finding rested on the court’s view that Google leveraged control in one layer of the stack to entrench its position in another — particularly through tying and related restrictions that limited publishers’ and rivals’ ability to route inventory in ways that would allow non Google tools to compete on the merits. The opinion also credited evidence that certain product and policy choices advantaged Google’s exchange and hindered rivals from achieving the scale and data access necessary to compete effectively, ultimately harming the competitive process on the publisher side.

The case now moves into its remedies phase, where the central question mirrors the challenge in the search case: How far should the court go to ‘untether’ interconnected pieces of the stack? Judge Brinkema will determine whether conduct remedies, structural relief, or a combination are warranted. Google has already stated that it plans to appeal aspects of the liability ruling, setting up a familiar narrative in major tech cases — a significant liability decision on paper, followed by a remedies battle and appellate review that will determine the practical impact, likely years down the line once the long legal process has finished playing out.

FTC v. Meta: The court treats ‘personal social networking’ as yesterday’s market

After a bench trial, Judge Boasberg ruled for Meta on 18 November 2025, rejecting the FTC’s monopolisation claims. The key dispute centred on the FTC’s proposed relevant market for ‘personal social networking services’ — essentially, apps used for sharing and keeping up to date with friends and family. The court was sceptical that this framing still reflected the competitive reality of how social media operates today. As Judge Boasberg noted, ‘how times have changed’, emphasising that the Facebook and Instagram of today ‘bear little resemblance’ to the platforms users or the industry would have recognised in the 2010s.

That evolution mattered because it undermined the FTC’s efforts to separate Meta’s apps from the broader attention-and-content competition that has increasingly come to define the space. The court concluded that ‘personal social networking’ no longer functioned as a distinct market and credited evidence showing that users readily substitute across services that blend friend and family sharing with interest based discovery and content consumption. In practical terms, the court found that platforms like TikTok and YouTube meaningfully constrain Meta. Once those constraints were recognised, the FTC failed to show that Meta currently holds monopoly power.

The ruling also underscores that a historical ‘buy or bury’ narrative does not relieve the government of proving present day monopoly power and competitive harm. Judge Boasberg repeatedly stressed that whatever Meta’s position may have been in the past, Section 2 liability turns on whether the company maintains monopoly power now in a market that has changed materially since the case was filed.

On 20 January 2026, the FTC filed a notice of appeal to the DC Circuit.

UK: The DMCC Act Takes Effect

The UK’s Digital Markets, Competition and Consumers Act (DMCC Act) came into force on 1 January 2025, introducing a bespoke, effects-based digital regulation regime. Unlike the EU’s DMA, it allows the CMA to impose tailored conduct requirements and, where necessary, pro-competitive interventions on firms designated with Strategic Market Status (SMS) — those with substantial and entrenched market power in strategically significant digital activities. SMS designations apply for five years.

In October 2025, the CMA issued its first SMS designations, identifying Google for search and search advertising and, subsequently, Google and Apple for their mobile platforms. With these designations now final, the CMA will move into implementation in 2026, consulting on conduct requirements and assessing potential interventions. Noncompliance carries significant exposure, including fines of up to 10% of global turnover and potential private damages actions.

The CMA also concluded its cloud services market investigation in July 2025, recommending Microsoft and AWS for potential SMS investigations, with decisions expected in early 2026.

As with the DMA, enforcement under the DMCC Act has prompted debate over pace and ambition. The CMA appears to be balancing political pressure to avoid deterring investment against the need to establish the regime’s credibility. Whether enforcement accelerates to match the EU’s increasingly assertive DMA approach — or remains more restrained, in line with recent UK merger control practice — will be a key issue in 2026.

The DMCC Act also introduces a two-track merger regime for technology firms. SMS-designated companies are subject to a parallel, mandatory and suspensory transaction reporting system alongside the UK’s voluntary regime, increasing procedural complexity and execution risk for global deals, particularly those involving minority investments, partnerships, or nontraditional acquisitions.

EU: Landmark New Phase of Enforcement for the DMA

Across the Channel, DMA enforcement moved into a decisive new phase in 2025. The European Commission shifted from monitoring and preliminary inquiries to active investigations and noncompliance decisions, imposing its first-ever DMA fines in April: €500 million on Apple for unlawful steering practices in breach of Article 5(4) DMA and €200 million on Meta for its ‘consent or pay’ model, contrary to Article 5(2) DMA.

In October, the Commission launched a call for tenders to assess the impact of emerging technologies — notably generative AI — ahead of the DMA’s first statutory review in May 2026, raising the prospect of extending the regime to additional services, potentially including AI.

These steps confirm that the DMA is an enforcement regime with real teeth. The Commission has shown its readiness to impose fines and behavioural remedies, making proactive compliance essential for gatekeepers and reinforcing the DMA as a viable tool for challenging anticompetitive conduct in digital markets.

At the same time, enforcement has met legal and political resistance. Apple is challenging the designation of the App Store and related interoperability obligations before the General Court, following ByteDance’s earlier unsuccessful appeal. The outcome, expected in 2026, will be critical in defining the scope of the Commission’s powers under the DMA. The litigation also highlights growing transatlantic tensions, with senior US officials criticising the DMA as an undue burden on US tech firms — raising questions as to how geopolitical pressures may shape enforcement going forward.

Labour Market Antitrust Practices Under Scrutiny

EU Issues First Fines for Labour Market Cartel

2025 marked a turning point in antitrust enforcement of labour markets, with clear implications for technology companies in talent-intensive sectors. In June, the European Commission imposed €329 million in fines on Delivery Hero and Glovo for exchanging sensitive information and engaging in no-poach and market-allocation arrangements — the Commission’s first labour market cartel decision and its first sanctioning of the anticompetitive use of a minority shareholding to facilitate collusion.

The decision sits within a broader global crackdown. In the US, DOJ and FTC guidance in January 2025 underscored potential civil and criminal liability for no-poach conduct. In the UK, the CMA issued its first labour market infringement decisions, fining the BBC, BT, IMG, and ITV £4.2 million for collusion on freelance pay, while publishing new guidance in September 2025 to reinforce its enforcement focus.

For the technology sector, where competition for AI, machine learning, and engineering talent is central, these developments are particularly acute. Regulators are increasingly treating hiring and wage restraints as hardcore infringements. Tech companies should review compliance frameworks, train HR and legal teams, and avoid formal or informal restrictions on recruitment. Only narrowly tailored, objectively necessary no-poach provisions ancillary to legitimate collaborations are likely to be defensible — and then only in exceptional cases.

2026 Previewed: What to Expect from the Year Ahead

2026 is likely to bring both continuity and recalibration. Scrutiny of Big Tech will remain intense, with regulators in the UK, EU, US, and beyond pressing ahead with investigations, designations, and enforcement that continue to test existing legal frameworks. At the same time, political headwinds, particularly transatlantic tensions over trade, investment, and technology policy, are unlikely to ease and will continue to shape digital regulation.

Businesses should expect sustained regulatory vigilance, especially in digital markets, AI, and other frontier technologies, alongside a measure of growing pragmatism in some jurisdictions as pressure mounts to support competitiveness and growth. In the United States, that pragmatism has translated into a more disciplined posture: continued willingness to litigate major conduct cases and challenge deals where the theory of harm is conventional and well supported, paired with greater caution about using antitrust to regulate emerging technology markets absent clear, traditional harms. Remedies — and the realism required to navigate them — will remain central as courts continue to favour relief closely tied to the exclusionary mechanisms actually proved.

For technology companies, this tension will define 2026: robust oversight of platform power and digital ecosystems, tempered by increasing sensitivity to innovation, scale, and economic impact. Outcomes will turn increasingly on deal narrative, jurisdictional sequencing, and the quality of regulatory engagement.