COMPETITION IN THE INFORMATION AGE

By James Mellsop, Mark Ponder, Veronica Postal, and John Scalf¹

Data is playing an increasingly central role in the modern economy. The large-scale collection and analysis of data has allowed firms to become more efficient while providing customers with improved product quality at lower prices. It has also allowed firms to tailor their product offerings to better meet the needs of disparate consumer segments. However, while acknowledging the procompetitive uses of this resource, antitrust authorities around the world have also raised concerns about its potential to harm competition. In this article, we examine three theories of harm that have emerged from recent antitrust enforcement in the United States: data as a strategic input for product development, the anticompetitive misuse of consumer data, and access to competitively sensitive data to unfairly disadvantage rivals. We explore the broader context of the economic principles underpinning these theories, and the related case law.

The digital revolution that has characterized the Information Age has resulted in a substantial share of economic activity moving online. This shift has been amplified by innovations in computer science and information technology over the past two decades, as mobile technology, more powerful computers, more sophisticated algorithms, and artificial intelligence ("AI") have become widely available and incorporated in many products adopted by consumers and firms. This transformation in the key locus of economic activity has been accompanied by an unprecedented expansion in the availability and usefulness of data. Many firms have started collecting and processing vast amounts of data, relying on the insights provided by these data to guide their decision making. Data affords firms a better understanding of what consumers demand, allowing firms to produce more desirable products at cheaper prices. However, concerns have been expressed that data can potentially also become an avenue to unfairly disadvantage rivals or a risk to consumer privacy. While the economic literature has not formed a clear consensus on how the unprecedented availability of data in the digital economy has impacted market structure or consumer welfare, regulators and policymakers have taken notice, and have started more closely scrutinizing how firms and digital platforms use data.

In his Executive Order on Promoting Competition in the American Economy, former U.S. President Joe Biden called on his Administration to "enforce the antitrust laws to meet the challenges posed by new industries and technologies, including the rise of large Internet platforms, especially as they stem from serial mergers, the acquisition of nascent competitors, the aggregation of data, unfair competition in attention markets, the surveillance of users, and the presence of network effects."² Furthermore, the executive order called on the U.S. Federal Trade Commission ("FTC") to address "persistent and recurrent practices that inhibit competition" by exercising its statutory rulemaking authority specifically with regard to "unfair data collection and surveillance practices

[Page 19]

that may damage competition, consumer autonomy, and consumer privacy."³ In line with this mandate, a review of enforcement actions by the U.S. Department of Justice ("DOJ") and FTC in recent years shows that the agencies have initiated a number of lawsuits and investigations that delved into the role of data in potentially granting firms an anticompetitive advantage or in harming consumers. While the Trump administration has signaled a move away from the Biden administration’s embrace of novel, expansive theories, it is still pursuing strong antitrust enforcement, especially in Big Tech and other key sectors, and is increasingly emphasizing the role of data, while relying on more traditional legal theories.

In this article, we examine recent antitrust enforcement investigations and lawsuits related to concerns over data. We explore three primary theories of harm considered by the agencies and unpack the economic principles underpinning these theories. Through this analysis, we aim to place recent regulatory perspectives within a taxonomy of theories of harm related to data.

I. DATA AS AN INPUT INTO PRODUCT DEVELOPMENT

The first concern that has been at the forefront of the antitrust analysis of data has been the importance of data as an input in product development. In the modern digital economy, data has emerged as a critical asset for businesses, often described as the "new oil" due to its value in creating competitive advantages.⁴ From an antitrust perspective, one concern regarding data has been the potential for the accumulation and control of large datasets to entrench established firms with market power, giving them a competitive edge which ultimately might lead to higher prices or a lower quality product.

From an economic perspective, data can provide significant competitive advantages in several ways.⁵ First, the process of collecting and processing large amounts of data often involves substantial fixed costs. As these costs are spread out over a growing customer base, the average cost per customer decreases. This means that new entrants in the market might need to operate with lower profits compared to established firms while they try to build their own customer base. The profitability advantage enjoyed by larger firms might allow them to refine their products, personalize services, and improve efficiency in ways that smaller firms or new entrants cannot easily replicate.⁶ Second, network effects in data-driven markets may exacerbate the competitive importance of data. In many digital markets—for example, social media or online marketplaces—the value of a service

[Page 20]

increases as more people use it. In turn, the more consumers use a service, the larger the volume of data that is available to the service provider.

The interaction of these two effects can lead to a data-driven advantage which creates a "feedback loop" where already successful firms become even more successful—more data leads to better services, attracting more users, and generating even more data. In some cases, the fixed costs associated with data collection and processing can be so high and the network effects so substantial that they become a barrier to entry, making it difficult for new players to enter the market. This dynamic can create a winner-takes-all scenario, where once a firm gains a significant market share, it is difficult for others to compete effectively.⁷

Moderating the potential for anticompetitive harm is the fact that network effects and scale economies can be welfare improving, as large firms may offer a superior product at a lower price. As such, it is typically not the collection and analysis of large datasets that leads to anticompetitive harm but instead the leveraging of a dominant market position to disadvantage rivals. Absent such conduct, the potential pro-competitive effects arising from data collection can be significant. Additionally, the competitive advantages conferred by data do not mean that access to data is always necessary to develop competitive products.

Data, unlike many other resources, is non-rivalrous—its use by one firm does not preclude its use by another. This characteristic potentially allows for multiple firms to compete if they have access to the same or similar data sets. Data may also be fungible in that different data sources may be used to solve the same problem. For instance, financial technology firms using alternative data in lieu of traditional credit reports in their scoring models when underwriting loans saw significant growth in recent years.⁸ Another moderating factor is that the value of data is not static but can depreciate rapidly. Today’s valuable dataset might be obsolete tomorrow. Thus, the competitive advantages afforded by data might not be as insurmountable as they appear, especially in fast-moving sectors where innovation can quickly change the competitive landscape.

Ultimately, the focus should be on the ability to generate insights from data rather than on the data itself. New entrants with superior algorithms or analytical capabilities could outcompete incumbents, even with access to smaller or less comprehensive datasets. Also, diminishing returns to data collection might limit the potential advantages obtained by incumbent firms. Statistical precision increases with the square root of the sample size, meaning that in order to improve an estimate by a factor 10 it is necessary to increase the sample size by a factor of 100.⁹ Similarly, machine learning algorithms frequently exhibit log-linear performance growth with respect to the number of samples used, meaning that

[Page 21]

exponentially more data is necessary to make modest improvements in accuracy.¹⁰ As such, large gains can frequently come from improved algorithms rather than through more data.

Concerns about data’s role as an input into product development have been central to the DOJ’s theory of harm in the recent case it has brought against Google regarding Google’s business practices in online search and search advertising.¹¹ The DOJ has alleged that Google, the world’s leading search engine, has maintained and extended its market power in search and search advertising by striking exclusive agreements to be the default search engine on a large portion of the major devices and services where users perform search. This includes mobile telephones, wireless carriers, and internet browsers. All told, the DOJ has alleged that Google’s default search engine agreements cover nearly 60 percent of all searches performed in the U.S. with an additional 20 percent covered by Google owned-and-operated properties like Chrome browser.¹²

While the DOJ has alleged that this has directly foreclosed competing search engines by steering default search engine traffic to Google, the DOJ has also theorized a dynamic effect due to the data Google has been able to harvest from its search engine users and that its rivals have been denied. The theory is that search queries create data, which can be used to improve search engine results, which drives further users to the search engine, creating even more data and leading to a strong product quality feedback cycle. As stated by the DOJ in its opening statement in the trial, data is "oxygen for a search engine" and the exclusive agreements Google has struck amounted to a "data fortress" that competitors cannot break through.¹³

Google, on the other hand, has argued that data is not nearly as central to search quality and that its marginal value to Google’s search algorithms diminishes with scale, implying that other competitors can create competing search engines with much lower volumes of data.¹⁴ As noted by Google’s Chief Economist in a 2009 interview, "it’s not the quantity or quality of the ingredients that make a difference, it’s the recipes."¹⁵ Despite these arguments, the court ruled in August 2024 that Google had violated Section 2 of the Sherman Act and that the use of these exclusive agreements "prevented rivals from accumulating the queries and associated data, or scale, to effectively compete." In September 2025, the court ordered Google to share "certain search index and user-interaction

[Page 22]

data" with competitors as part of a remedy.¹⁶ Additionally, the Court also discussed the rapid rise of generative AI ("GenAI") during the period between the liability trial and the remedies hearing, emphasizing the "nascent competitive threat" that GenAI products pose to general search engines. This discussion reinforces the notion that innovation can quickly reshape the competitive landscape and diminish advantages derived from data.

The lawsuit brought by the FTC against Amazon in September 2023 is relevant to the proposition described above that the collection of consumer data can lead to network effects and economies of scale but might not alone constitute harm to competition.¹⁷ Instead, harm to competition might arise if a firm leverages its comparative advantage to prevent competitors from achieving the necessary scale to compete effectively. The FTC claims that access to shopper data is one of the sources of network effects for "online superstores" like Amazon.¹⁸ The complaint alleges that online superstores use data to personalize their users’ shopping experiences, and that as an online superstore gains more scale, this effect becomes more powerful.¹⁹ Thus, despite the obvious benefits that personalization and tailored shopping experiences bring to consumers, the FTC alleges that data serves as a barrier to entry since "[p]rospective entrants would have to acquire a sufficient shopper base to obtain enough data to offer this level of personalization."²⁰

The FTC contends that Amazon’s access to considerable amounts of consumer data allows it to personalize the shopping experience of its customers and to offer a service that is unmatched by competitors not able to achieve the requisite scale. Because these competitors cannot offer a comparable shopping experience, they would normally attempt to compete with Amazon through lower prices. That is, when consumers care about the price and quality of their shopping experience, and competing marketplaces cannot offer a similar quality experience, then these competitors must offer lower prices to attract customers. This strategy would allow them to start building a consumer base and eventually allow them to offer a service of comparable quality. Thus, over time competition would tend to increase the quality of the shopping experience and lead to lower prices.

However, the FTC has alleged that Amazon’s contracts and algorithms preclude vendors from selling at lower prices on competing online marketplaces, in turn foreclosing a key avenue of competition that could be used by other firms to obtain scale. As such, the alleged harm to competition does not arise from Amazon’s data collection. Rather, data create an alleged barrier to entry that competitors need to overcome in order to succeed, but Amazon’s efforts to prevent vendors from offering lower prices on other platforms deprived competitors of the ability to do so.

[Page 23]

Importantly, not all barriers to entry and expansion are due to economies of scale and network effects. The data generated by consumers when using a firm’s product or service may also be leveraged by the firm to increase switching costs and create customer lock-in. In situations where the data are not fungible, the firm may hinder portability or restrict the access of other firms to key data in order to stifle competition. Consider, for instance, the allegations in the recent class action lawsuit Gamboa v. Apple.²¹ In this case, the plaintiffs alleged that Apple "leveraged its power in the mobile device markets to exclude competition in the market for Full-Service Cloud Storage."

Specifically, the plaintiffs alleged that Apple prevented third-party cloud storage competitors from accessing "Restricted Files" associated with device settings, apps, and app data, which had the effect of limiting a customer’s ability to back up their mobile device on third-party cloud storage platform. According to the plaintiffs, this restriction in effect created an illegal tying arrangement where a customer with an Apple mobile device needed to purchase Apple’s cloud storage option, iCloud, if they wanted access to Full-Service Cloud Storage. The complaint further asserts that because customers placed a high value on full-service functionality, the ultimate impact was to foreclose competition and that "consumer choice is gravely diminished." The case is still ongoing, with the Court recently denying Apple’s motion to dismiss in June 2025, stating that "Plaintiff’s plausibly allege their monopoly claim . . . based on allegations of tying."²² This lawsuit underscores the potential for consumer data to be used to increase switching costs and insulate a dominant firm from competitive pressures.

The highlighted cases underscore the role of data in digital market competition and highlight the complex interplay between data control, market power, and antitrust enforcement in a rapidly evolving digital landscape.

II. ANTICOMPETITIVE MISUSE OF PERSONAL DATA

Another competitive concern that has been cited in recent antitrust enforcement is that, similarly to how market power can be exercised by firms to increase prices, market power may also be leveraged to force greater disclosure of personal information from consumers.²³ On top of that, whether a firm has extracted "too much" data is unclear. The economic literature has not formed a consensus on how to systematically assess optimal privacy. The balance between consumer privacy and information sharing between consumers and firms is not generally defined. Consumer privacy differs from typical economic goods as its value is often context dependent, privacy sensitivities and preferences vary by individual, and because what constitutes sensitive information differs

[Page 24]

across individuals.²⁴ Because data is non-rivalrous, the socially optimal level of consumer privacy might be higher than what is individually optimal for consumers.²⁵ Non-rivalry means that the value of data is not diminished when it is used by multiple firms, and thus sharing data broadly across firms might lead to large social gains, if at the cost of a reduction in the privacy enjoyed by individual consumers. Moreover, situations where sharing personal information can reveal information about other individuals might result in excessive data sharing compared to the social optimum.²⁶

The protection of consumer privacy has historically fallen outside the scope of antitrust enforcement, which traditionally has focused on conduct that resulted in higher prices, lower output, or lower quality for consumers, and instead fell under the mandate of consumer protection agencies. However, with the rise of the digital economy, the distinction between antitrust and consumer protection has been blurred. Firms are not only able to collect, store, and analyze consumer data on an unprecedented scale to inform internal business decisions, but they can also commercialize data for purposes that may or may not be aligned with the interests of consumers. The use of behavioral marketing in digital advertising has become more ubiquitous, fueled by the availability of consumer data on the background and preferences of consumers.²⁷

The revenue generated by digital advertising has allowed firms to provide many digital services to consumers at no monetary cost, such as social media and music streaming. While this business model has the potential to generate high levels of consumer surplus, one concern is that personal data has become a resource that can be monetized, reducing the overall level of consumer privacy afforded to consumers. The theory is that consumers, often unknowingly or without viable alternatives, pay for the services with their personal information, which becomes less protected and more susceptible to misuse when a firm with market power can set the terms of user engagement.

However, it must be noted that such business models can generate substantial benefits for consumers. The collection and analysis of consumer data can drive innovation. This data enables firms to improve existing products, develop new services, and tailor offerings to individual user preferences. Moreover, the business model of providing digital services for free, funded by advertising that relies on consumer data, is an arrangement that allows a broader section of the population to access these services.

A concern about the anticompetitive misuse of personal data was expressed in the FTC’s recent lawsuit against Facebook’s parent company, Meta. Facebook operates a two-sided platform where consumers have access to social networking services for free. On the

[Page 25]

other side of the platform, Facebook monetizes its product through "social advertising," a type of digital marketing that allows businesses to deliver personalized advertisements to consumers. Central to this business model is Facebook’s collection and distribution of data via its API systems. In this lawsuit, the FTC has alleged that Facebook has conditioned access to its API systems to anticompetitively exclude rival personal social networking providers and has unlawfully maintained a monopoly in personal social networking.²⁸ Facebook earns a considerable share of its revenue by allowing businesses to place personalized advertisements on its platforms, and the FTC has alleged that Facebook has harmed competition and consumers in a market for social advertising. In addition, the FTC has alleged harm directly to the users of its social networking service: "Facebook has been able to provide lower levels of service quality on privacy and data protection than it would have to provide in a competitive market."²⁹ Specifically, the FTC has alleged that Facebook excluded nascent competitors, preventing them from becoming competitive constraints that would have resulted in a better level of service offered to users, including through improved "availability, quality, and variety of data protection privacy options for users, including, but not limited to, options regarding data gathering and data usage practices."³⁰ While the case is still ongoing, it showcases the FTC’s concern with the way in which competition may impact firms’ policies toward the protection of user privacy.

Two acquisitions by Amazon have similarly been under scrutiny for concerns related to the potential use of sensitive consumer data. U.S. Senator Elizabeth Warren and five U.S. Representatives petitioned the FTC to oppose Amazon’s proposed acquisition of iRobot, a smart vacuum manufacturer, citing "Amazon’s record of infringing on consumers’ privacy" and concerns with Amazon’s "growing surveillance powers" associated with iRobot’s mapping technology.³¹ While the FTC has not publicly released the details of its investigation into the iRobot deal, the Competition and Markets Authority ("CMA") investigated the privacy implications of the deal, and Amazon’s potential ability to collect and combine various sources of consumer data, as one potential area of concern.³²

[Page 26]

Similar concerns were raised with respect to Amazon’s acquisition of One Medical, a membership-based primary care provider. After the acquisition was completed, the parties released a statement assuring that "nothing about this acquisition changes Amazon or One Medical’s commitment to privacy or the strong protections we have for Protected Health Information"³³ However, while certain individual health information is protected by HIPAA, the law does not preclude the use of anonymized, aggregated data or of data that consumers agree to share with third parties. While the use of such data could improve consumer welfare by allowing Amazon to further tailor its product offerings to the specific needs of its customers, the FTC expressed concerns with respect to how the merging parties would handle potentially sensitive consumer information. While the FTC ultimately did not challenge the acquisition, it released a joint statement stating that "the parties and the market more broadly should be on notice that the Commission will continue to monitor this space and bring enforcement actions whenever the facts warrant."³⁴ Again, these investigations demonstrate the increasing focus of enforcement agencies on privacy and the use of sensitive consumer data as a potential avenue of consumer harm.

III. ACCESS TO COMPETITIVELY SENSITITIVE DATA TO UNFAIRLY DISADVANTAGE RIVALS

Still another competitive concern regarding data pertains to situations where one company may gain access to a large amount of sensitive data on competitors’ operations, for instance through their importance as a service provider to other competitors in the industry or through a third-party provider. The concern is that access to competitively sensitive information may lead to reduced competitive constraints in an industry, either by increasing the ability of a firm to disadvantage its rivals or by increasing the likelihood of coordinated effects.

For example, FTC Chair Lina Khan indicated such concerns about Amazon, arguing that Amazon has used data on its third-party sellers’ sales on Amazon’s marketplace to "unfairly" compete with these sellers.³⁵ This includes identifying popular third-party products and launching competing versions of these products under its private label, which Chair Khan claims allows Amazon to increase sales without incurring the risk incumbent third-party sellers face when launching a product. ³⁶ Chair Khan argues that Amazon can engage in such actions because of its dominance as a sales platform that third-party sellers must access even though they may have concerns with respect to Amazon’s use of their

[Page 27]

data.³⁷ However, as acknowledged by Chair Kahn in her academic work, these practices are not limited to Amazon, and brick-and-mortar retailers have also used similar data, including other brands’ sales records, to introduce private labels.³⁸

The potential for the sharing of competitively sensitive information leading to coordinated effects was a key argument advanced in the FTC’s recent complaint against Agri Stats in September 2023.³⁹ Agri Stats aggregates and distributes operational data from the largest meat producers in the United States, including aggregate information on prices, wages, costs, inventories, and production plans. In its complaint, the FTC has alleged that "Agri Stats operates an information-sharing scheme that allows processors to exchange competitively sensitive information about their operations and sales that is comprehensive, granular, current, and available exclusively to processors"⁴⁰ and that "[w]ith Agri Stats’ encouragement and facilitation, Agri Stats’ processor subscribers use the information collected and distributed by Agri Stats to increase and stabilize prices and reduce the supply of meat."⁴¹

Economic theory suggests that firms may be more likely to sustain collusive behavior if deviations from the supra-competitive price can be readily identified and punished. If deviations were difficult to discern, the firms may have an incentive to charge a price below their competitors to gain additional business and higher gross profits. The FTC has alleged that by sharing competitively sensitive information, "Agri Stats reduces common challenges to coordination—distrust among competitors, and ‘cheating’ on agreements."⁴² Of course, informational exchanges can also have procompetitive effects, such as allowing companies to benchmark their performance to increase efficiency, reduce costs, or improve demand forecasting.⁴³ Nevertheless, this case demonstrates how regulatory agencies continue to scrutinize whether the sharing of potentially competitively sensitive information may harm competition in a broad sense, rather than having a narrow focus on cases in which specific competitors were disadvantaged.

Similar concerns were echoed in recent cases involving algorithmic collusion. For example, in its complaint against RealPage, the DOJ alleged that the company’s algorithmic rent-setting software exploited sensitive data provided by subscribing landlords to coordinate and maximize rent prices.⁴⁴ Economic theory suggests that the

[Page 28]

use of pricing algorithms can lead to procompetitive outcomes, such as prices that are more responsive to market conditions and a more efficient allocation of resources. The DOJ’s allegations center around whether RealPage’s software collects and analyzes non-public pricing data, such as rental applications, executed new leases, renewal offers and acceptances, and forward-looking occupancy information, to generate recommendations for rental prices which in effect allowed competing landlords to coordinate their rental prices.⁴⁵ This competitively sensitive information is alleged to be more granular, timely, and comprehensive than alternatives available to potential renters or competitors that do not subscribe to RealPage’s services.⁴⁶

According to the DOJ, "[c]ompetitively sensitive data collected from competing landlords is a critical input to RealPage’s revenue management software" and "[n]o other revenue management company can match RealPage’s access to landlords’ nonpublic, competitively sensitive rental data."⁴⁷ These allegations raise fundamental questions about balancing the pro-competitive benefits of algorithmic pricing with the need to safeguard competition by preventing the sharing of competitively sensitive information. They also challenge the definition of what constitutes competitively sensitive data and whether the availability of modern data analysis tools can blur the lines between legitimate market strategies and anti-competitive behavior. As stakeholders in the housing market grapple with these developments, it is crucial to scrutinize how algorithmic pricing tools can be designed to support fair competition without facilitating collusion.

There are procompetitive reasons for industry leaders to also serve as platforms and service providers even if it can lead to the exchange of competitively sensitive information. There may be synergies that drive industry-specific solutions to be developed by industry participants, instead of third parties not involved in the industry. For example, Visa, the payment card company, was founded by Bank of America as an internal credit card product and was later licensed by Bank of America to other banks once the system proved successful internally. In other words, it is not surprising that an industry solution would be developed by a current industry participant. There also may be economies of scale that drive an industry to a centralized solution, even if it involves access to competitively sensitive information. For example, in the case of Amazon, there are clear efficiencies to centralizing the functions of building an e-commerce website. These efficiencies can ultimately benefit consumers through an expansion in the variety of available products and through a reduction in price. Moreover, the fear of appropriation of competitively sensitive data can be mitigated by the use of firewalls, which are common in many industries.⁴⁸

The question of whether a firm’s access to a competitor’s data will give the firm an unfair advantage might also be affected by the competitive constraints faced by the firm. If

[Page 29]

there are multiple companies competing to provide the same service, then the firm would likely have an incentive to implement safeguards for consumer data or firewalls or come to other agreements regarding the access to competitively sensitive data.

This dynamic featured prominently in the FTC’s action to block the proposed acquisition of Aerojet Rocketdyne ("Aerojet") by defense contractor Lockheed Martin.⁴⁹ Aerojet is one of two companies supplying solid rocket motors ("SRM") to major defense contractors (also known as primes) such as Lockheed Martin and may supply these parts to multiple primes competing for the same opportunity. As such, Aerojet allegedly has access to competitively sensitive information from various prime contractors about their system designs. An important feature of this market underlying the FTC’s theory of harm was the limited supply of SRM manufacturers, which was perceived to lessen the incentive to safeguard the competitively sensitive information. In challenging the transaction, the FTC alleged that the deal "heightens the incentives for Lockheed to misuse . . . competitively sensitive, non-public information of rival primes" in order to gain a competitive advantage for a given proposal.

Further, the FTC has alleged that Lockheed Martin, being one of the largest purchasers of SRMs, would "be privy to competitively sensitive, non-public information relating to Aerojet’s only SRM rival, Northrop [Grumman]" and would be able to use this information to advantage its newly acquired SRM business. In its complaint, the FTC asserted that "[p]reventing such potential anticompetitive exchanges of information is necessary to maintain effective competition in the Relevant Markets to ensure that innovation, price, and/or performance for these important U.S. military systems is not negatively impacted." Facing resistance from the FTC, the parties eventually decided to abandon the transaction. This case demonstrates that the FTC not only considered the potential for a vertically integrated company to use an informational advantage to lessen downstream competitive constraints, but also using said information to advantage its upstream business.

The DOJ recently raised similar concerns in its challenge to the merger of UnitedHealth Group ("UHG") and Change Healthcare ("Change").⁵⁰ UHG, the largest health insurer in the United States, proposed to acquire Change, a health technology company used by a large share of health insurers and healthcare providers primarily in the processing of healthcare claims. UHG also had a smaller presence in claims processing as part of its Optum business. The DOJ in its complaint alleged that the merger would allow UHG to access rivals’ claims data via their business with Change, which would allow UHG to "slow its rivals’ innovations, reverse-engineer its rivals’ proprietary plan and payment rules, preempt their competitive strategies, and compete less vigorously for certain customers," which "would prove profitable to United while harming competition."⁵¹

[Page 30]

The court reasoned, however, that UHG through Optum already had access to much of the competitively sensitive information that the government sought to prevent UHG from gaining access to as part of the deal and that the government did not provide any evidence as to what the incremental value of the additional data that UHG would have access to via the Change acquisition.⁵² Moreover, the court found that UHG’s Optum subsidiary has strong incentives to protect its customers’ data since those customers expect that their data will be kept out of the hands of UHG’s insurance business. In fact, UHG already has a system of firewalls in place to protect Optum’s data from being used by UHG’s insurance business. The court found that the government put forward no argument as to why these incentives would change with the Change acquisition. The court’s finding in the challenge of the UHG/Change merger highlighted the importance of real-world evidence in examining whether a theory of harm centered around data sharing is plausible. As the court noted, "antitrust theory and speculation cannot trump facts."⁵³

The DOJ’s recent suit against Google for monopolization of multiple online advertising technology products also raises antitrust concerns about alleged access to competitively sensitive information.⁵⁴ In this case, the DOJ has alleged that Google has used its market power in the provision of publisher ad servers to harness a trove of historical information on the bids publishers received for their ad inventory to create a proprietary algorithm called "Smart Bidding" that would predict the bids publishers would receive for their future ad inventory. Using this algorithm, the DOJ claims that Google would gain an "insider" advantage in the auction for publishers’ ad inventory.⁵⁵

In its April 2025 ruling, the Court held that Google was liable for both monopolization and tying. In their opinion, the judge specifically cited the "increasingly detailed knowledge about the billions of people who have used its products, including by collecting data pertaining to their web browsing, search activity, physical location, demographic characteristics, app usage, communications, shopping activity, and device and network information" that Google has collected over the past two decades as giving it a "data advantage" over its competitors.⁵⁶ This decision is one of the first explicitly mentioning data as a source of competitive advantage, and it could pave the way for changes in how tech companies operate and how they handle user data moving forward.

IV. CONCLUSION

In this era of digital proliferation, data has not only become more ubiquitous but also potentially more valuable. This surge in data availability, while largely beneficial for consumers, also has raised concerns with the antitrust agencies about its potential for reinforcing market positions and for creating new avenues for consumer detriment. The

[Page 31]

contemporary antitrust narrative is increasingly dominated by three concerns about data: its importance to product development, the potential misuse of consumer data, and the unfair exploitation of competitively sensitive information. These concerns highlight a need for antitrust frameworks that balance the dual objectives of fostering innovation and safeguarding consumer interests against the backdrop of increasingly data-centric markets.

[Page 32]

——–

Notes:

^1. James Mellsop is a Senior Managing Director and the Chair of NERA’s Global Antitrust and Competition Practice. Dr. John Scalf is a Director in NERA’s Antitrust and Intellectual Property Practices. Dr. Mark Ponder and Dr. Veronica Postal are Senior Consultants in NERA’s Antitrust and Intellectual Property Practices. The opinions expressed are those of the authors and do not necessarily reflect the views of NERA Economic Consulting or other NERA experts.

^2. Exec. Order No. 14036, 86 Fed. Reg. 36,987 (July 14, 2021).

^3. Id.

^4. See, e.g., The world’s most valuable resource is no longer oil, but data: The data economy demands a new approach to antitrust rules, Economist (May 6, 2017), https://www.economist.com/leaders/2017/05/06/the-worlds-most-valuable-resource-is-no-longer-oil-but-data.

^5. Daniel L. Rubinfeld & Michal S. Gal, Access Barriers to Big Data, 59 ARIZ. L. REV. 339, 339-381 (2017), https://arizonalawreview.org/pdf/59-2/59arizlrev339.pdf; Andres V. Lerner, The Role of ‘Big Data’ in Online Platform Competition (Aug. 26, 2014), https://ssrn.com/abstract=2482780; Data, Digital Markets and Refusal to Supply, Ofcom economic discussion paper series, issue no. 6 (Dec. 7, 2022), https://www.ofcom.org.uk/__data/assets/pdf_file/0028/248950/Data,-Digital-Markets-and-Refusal-to-Supply.pdf.

^6. See, e.g., OECD Competition Committee, Barriers to Entry, Series Roundtables on Competition Policy, No. 58 (March 6, 2006), https://www.oecd.org/daf/competition/36344429.pdf.

^7. Id. at 33.

^8. Terri Bradford, "Give Me Some Credit!": Using Alternative Data to Expand Credit Access, Federal Reserve Bank of Kansas City, June 28, 2023, https://www.kansascityfed.org/Payments%20Systems%20Research%20Briefings/documents/9638/PaymentsSystemResearchBriefing23Bradford0628.pdf.

^9. See, e.g., Fumio Hayashi, Econometrics (Princeton University Press 2000).

^10. See, e.g., Hal Varian, Artificial intelligence, economics, and industrial organization, in THE ECONOMICS OF ARTIFICIAL INTELLIGENCE: AN AGENDA 399 (Nat’l Bureau of Econ. Res. Conf. Rep. Series, University of Chicago Press 2019); Vishaal Udandarao et al., No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance (April 4, 2024), https://arxiv.org/abs/2404.04125.

^11. Complaint, U.S. Dep’t of Justice v. Google LLC, No. 1:20-cv-03010 (D.D.C. Oct. 20, 2020), https://www.justice.gov/opa/press-release/file/1328941/download.

^12. Id.

^13. Steve Lohr, A Key Question in Google’s Trial: How Formidable Is Its Data Advantage? N.Y. TIMES, Sept. 18, 2023, https://www.nytimes.com/2023/09/18/business/google-antitrust-trial-data.html.

^14. Defendant Google LLC’s Pre-Trial Brief (Redacted), U.S. Dep’t of Justice v. Google LLC, No. 1:20-cv-03010-APM (D.D.C. Sept. 8, 2023).

^15. Tom Krazit, Google’s Varian: Search scale is ‘bogus,’ CNET (Aug. 14, 2009), https://www.cnet.com/culture/googles-varian-search-scale-is-bogus/.

^16. Memorandum Opinion, United States of America et al. v. Google LLC, No. 20-cv-3010 (APM) (D.D.C. Sep. 2, 2025)

^17. Complaint (Redacted), Federal Trade Commission v. Amazon.com, Inc., No. 2:23-cv-01495-JHC (W.D. Wash. Nov. 2, 2023), https://www.ftc.gov/system/files/ftc_gov/pdf/1910134amazonecommercecomplaintrevisedredactions.pdf.

^18. Id. at ¶ 180.

^19. Id. at ¶ 180.

^20. Id. at ¶ 180.

^21. Second Amended Class Action Complaint, Gamboa v. Apple, Inc., No. 5:24-cv-01270-EKL (N.D. Cali. Mar. 21, 2025)

^22. Order Denying Motion to Dismiss the Second Amended Complaint, Gamboa v. Apple, Inc., No. 5:24-cv-01270-EKL (N.D. Cali. June. 16, 2025)

^23. See, e.g., Nathan Newman, Search, Antitrust, and the Economics of the Control of User Data, 31 YALE J. REG. 401 (2014), https://openyls.law.yale.edu/bitstream/handle/20.500.13051/8199/14_31YaleJonReg401_2014_.pdf?sequence=2&isAllowed=y; Frank Pasqual, Privacy, Antitrust, and Power, 20 GEO. MASON L. REV. 1009 (2013), https://heinonline.org/HOL/Page?handle=hein.journals/gmlr20&div=37&g_sent=1&casa_token=&collection=journals.

^24. Alessandro Acquisti, Curtis Taylor, & Liad Wagman, 54 The Economics of Privacy, J. OF ECON. LITERATURE 442, 446 (2016). See also, Daniel J. Solove, The Myth of the Privacy Paradox, 89 GEO. WASH. L. REV.1 (2021).

^25. Charles I. Jones & Christopher Tonetti, Nonrivalry and the Economics of Data, 110 AM. ECON. REV. 2819 (2020).

^26. Daron Acemoglu et al., Too much data: Prices and inefficiencies in data markets, 14 AM. ECON. J.: MICROECONOMICS 218 (2022).

^27. See, e.g., OECD Directorate for Financial and Enterprise Affairs Competition Committee, Consumer data rights and competition—Note by the United States (June 12, 2020), https://www.justice.gov/atr/page/file/1316691/dl?inline.

^28. First Amended Complaint for Injunctive and Other Equitable Relief (Redacted), Federal Trade Commission v. Facebook, Inc., No. 1:20-cv-03590-JEB (D.D.C. Aug. 19, 2021), https://www.ftc.gov/system/files/documents/cases/ecf_75-1_ftc_v_facebook_public_redacted_fac.pdf.

^29. Id. at ¶ 221.

^30. Id. at ¶ 220.

^31. Letter from Senator Elizabeth Warren et al. to the Honorable Lina Khan, Chair, Federal Trade Commission (Sept. 28, 2022), https://www.warren.senate.gov/imo/media/doc/Letter%20to%20FTC%20on%20Amazon%20-%20iRobot%20Merger.pdf.

^32. While the CMA cleared the deal, the European Commission objected to the proposed acquisitions. In January 2024, Amazon and iRobot announced that they agreed to terminate the pending acquisition, before the FTC could not conclude its investigation. See Competition & Markets Authority, Anticipated acquisition by Amazon.com, Inc of iRobot Corporation: Decision on relevant merger situation and substantial lessening of competition, ME/7012/22 (June 16, 2023), https://assets.publishing.service.gov.uk/media/64be31862059dc000d5d2851/Amazon_iRobot_decision_-_NON-CONFIDENTIAL_-_FOR_PUBLICATION.pdf; European Commission, Commission sends Amazon Statement of Objections over proposed acquisition of iRobot (Nov. 27, 2023), https://ec.europa.eu/commission/presscorner/detail/en/IP_23_5990; Press Release, iRobot, Amazon and iRobot agree to terminate pending acquisition (Jan. 29, 2024), https://media.irobot.com/2024-01-29-Amazon-and-iRobot-agree-to-terminate-pending-acquisition.

^33. One Medical Joins Amazon to Make It Easier for People to Get and Stay Healthier, ONE MEDICAL (Feb. 22, 2023), https://www.onemedical.com/mediacenter/one-medical-joins-amazon/. See also, Rebecca Klar, Klobuchar asks FTC to investigate Amazon’s $3.9 billion move to acquire One Medical, THE HILL (July 22, 2022), https://thehill.com/policy/technology/3570452-klobuchar-asks-ftc-to-investigate-amazons-3-9-billion-move-to-acquire-one-medical/.

^34. Joint Statement of Chair Khan, Commissioner Slaughter, Commissioner Wilson, and Commissioner Bedoya Regarding Amazon.com, Inc.’s Acquisition of 1Life Healthcare, Inc., U.S. Federal Trade Commission (Feb. 27, 2023), https://www.ftc.gov/system/files/ftc_gov/pdf/2210191amazononemedicalkhanslaughterwilsonbedoya.pdf.

^35. Lina M. Khan, Amazon’s Antitrust Paradox, 126 YALE L.J. 710, 710-805 (2017), https://www.yalelawjournal.org/note/amazons-antitrust-paradox.

^36. Id. at 780-783.

^37. Id. at 780-783. Interestingly, these concerns are not reflected in the lawsuit brought by the FTC against Amazon in September 2023. See Complaint [Public Redacted Version], Federal Trade Commission v. Amazon.com, Inc., No. 2:23-cv-01495-JHC (W.D. Wash. Nov. 2, 2023), https://www.ftc.gov/system/files/ftc_gov/pdf/1910134amazonecommercecomplaintrevisedredactions.pdf.

^38. Khan, supra note 32, at 782.

^39. Complaint, U.S. Dep’t of Justice v. Agri Stats, Inc., No. 0:23-cv-03009 (D. Minn. Sept. 28, 2023), https://www.justice.gov/d9/2023-10/416782.pdf.

^40. Id. at 16.

^41. Id. at 51.

^42. Id. at 19.

^43. OECD Competition Committee, Information Exchanges between Competitors under Competition Law, Series Roundtables on Competition Policy, No. 115 (2010), https://www.oecd.org/daf/competition/48379006.pdf.

^44. Complaint, U.S. Dep’t of Justice v. RealPage, Inc, No. 1:24-cv-00710 (M.D.N.C. Aug. 23, 2024), https://www.justice.gov/archives/opa/media/1383316/dl?inline.

^45. Id. at 5, 15, 75.

^46. Id. at 17.

^47. Id. at 11, 15.

^48. For example, firewalls are commonly employed in investment banks between the brokerage and corporate advisory businesses so that insider information about corporate advice is not used to advantage the bank’s brokerage clients. Likewise, in government procurement, firewalls are often put in place so that a firm that already has a contract with a government is not unfairly advantaged in subsequent bids for new contracts.

^49. Complaint (Redacted), In re Lockheed Martin Corp. and Aerojet Rocketdyne Holdings, Inc., No. 9405 (F.T.C. Jan. 25, 2022), https://www.ftc.gov/system/files/documents/cases/d09405lockheedaerojetp3complaintpublic.pdf.

^50. Complaint, U.S. Dep’t of Justice v. UnitedHealth Group Incorporated, No. 1:22-cv-00481 (D.D.C. Feb. 24, 2022), https://www.justice.gov/media/1208936/dl?inline.

^51. Id. at ¶ 12.

^52. Memorandum Opinion, United States of America v. UnitedHealth Group, No. 1:22-cv-00481-CJN (D.D.C. Sept. 21, 2022), https://www.justice.gov/d9/2023-08/415418.pdf, at 35-36.

^53. Id. at 12, 33, 52.

^54. Complaint, United States of America v. Google LLC, No. 1:23-cv-00108 (E.D. Va. Jan. 24, 2023), https://www.justice.gov/atr/case-document/file/1566706/dl?inline.

^55. Id. at ¶¶ 256-261.

^56. United States v. Google LLC, No. 1:23-CV-108 (LMB/JFA), 2025 WL 1132012 (E.D. Va. Apr. 17, 2025).