Policy Analysis

Addressing Civil Investigative Demand Overreach at the Consumer Financial Protection Bureau

The Consumer Financial Protection Bureau has a history of issuing poorly justified and overbroad civil investigative demands. Statutory reforms should curb overbroad and burdensome demands, ensuring that the Bureau’s targets are treated more fairly.

May 20, 2026 • Policy Analysis No. 1018

By Solveig Singleton

The Consumer Financial Protection Bureau (CFPB, or the Bureau) often issues civil investigative demands (CIDs), a type of administrative subpoena that compels businesses and individuals to produce evidence such as documents and testimony. Many observers criticize the Bureau’s CIDs for being poorly justified, overbroad, vague, and costly.

Modern Supreme Court precedents leave CID recipients with few constitutional protections. Courts defer to the Bureau in legal challenges to CIDs, rendering constitutional and statutory limits on the Bureau’s powers toothless.

Legislators are considering reform proposals to bring accountability to the Bureau’s CID practice. The Civil Investigative Demand Reform Act of 2025 would curb overbroad CIDs by requiring them to make specific reference to particular facts, empowering recipients to petition for relief from burdensome or unnecessary demands, and allowing petitioners to appeal directly to federal court upon the denial of a petition.

Policymakers should consider additional reforms. For example, the Bureau could be required to prove that CID compliance costs are proportional to the seriousness of the violation. Reforms should ensure that petitions are decided not by the Bureau’s director but by neutral arbiters such as administrative law judges housed outside the agency.

The Bureau’s CID overreach can destroy lawful enterprises without corresponding benefits to consumers. Procedural reforms to the CID process are an important step. But ultimately, protection of firms’ and individuals’ liberty will require legislators to take a skeptical view of inessential substantive rules more broadly.

Introduction

Imagine you receive a notice telling you that your business is under investigation for unspecified violations of a dozen federal statutes. You and your staff spend hours compiling documents and answering questions and tens of thousands of dollars on attorneys. Months pass before you find out what, if anything, you have done wrong. Eventually, as your focus shifts from customers and compliance costs eat into revenues, your business fails.

The CFPB often pursues such troubling investigations using a type of administrative subpoena called a civil investigative demand (CID), which requires people and organizations to provide evidence relevant to an investigation.¹ Many observers have criticized the Bureau’s CIDs for being poorly justified, overbroad, vague, and costly; for example, the Bureau might proceed with extensive demands for documents without articulating grounds to believe the CID recipient violated any law.²

Under the Trump administration, the CFPB’s mission, staffing, and funding have been curtailed. Still, these changes leave the Bureau’s statutory subpoena powers untouched. This paper examines proposals to make lasting changes to curb the CFPB’s CID overreach.

Foundations of the CFPB’s CID Authority

Under 12 U.S.C. Section 5562, the CFPB may issue a CID to any person it “has reason to believe” may have information “relevant to a violation.”³ The Bureau may demand documents, tangible things, reports, and oral testimony.

Section 5562 requires the CID to “state the nature of the conduct constituting the alleged violation” and “the provision of law applicable to such violation.”⁴ CIDs should describe the documents the Bureau seeks with such “definiteness and certainty as to permit such material to be fairly identified.”⁵

Recipients may petition to have the CID modified or set aside based on constitutional or legal arguments. Petitions are decided by the Bureau’s director. CFPB rules determine whether demands and petitions remain confidential.⁶

Modern Supreme Court precedents set few limits on administrative subpoenas such as CIDS, but the Constitution does or should provide some limits in some cases.⁷ Essentially, the Fifth Amendment applies when the person whose information is sought by the government must do something, such as provide documents in response to a subpoena. The Fourth Amendment applies when the person whose information is sought need do nothing, because the government is conducting a search.⁸ Sometimes both amendments apply, but in other cases neither does, or the court’s application lacks rigor. This incomplete coverage means that agencies face little accountability for administrative subpoena overreach.

Generally, the Fifth Amendment’s right against self-incrimination limits the government’s subpoena power, stating that “[n]o person … shall be compelled in any criminal case to be a witness against himself.”⁹ The Fifth Amendment applies in civil cases if the evidence might be used in a criminal case, or if the penalty is punitive.¹⁰ The modern Supreme Court, however, holds that the Fifth Amendment does not cover voluntarily created records, including business records, and that collective entities such as corporations do not have Fifth Amendment rights.¹¹ The modern Court also holds that the Fifth Amendment protects only “testimonial” information, that is, relating to an assertion of fact.¹² The modern Supreme Court extends greater Fifth Amendment protection to oral testimony than to the contents of preexisting documents.¹³

When these controversial limitations are taken together, most Fifth Amendment challenges to the Bureau’s CID by a targeted business would fail.¹⁴ Most would involve the surrender of business records of a collective entity such as a corporation, even if the information the Bureau requested were testimonial.¹⁵

Under the Fourth Amendment, investigators conducting a search must show probable cause to obtain a warrant “particularly describing the place to be searched, and the persons or things to be seized.”¹⁶ However, the Supreme Court treats administrative subpoenas as less intrusive than a physical search: A subpoena is a “constructive search.”¹⁷ Thus, the modern Court requires only that administrative subpoenas be authorized by statute and be reasonable.¹⁸ That is, an administrative subpoena must be limited in scope, reasonably relevant, and not unreasonably burdensome.¹⁹ The subpoena must not be issued in bad faith, and the recipient must have an opportunity for review before incurring penalties for noncompliance.²⁰

The Court’s already undemanding standards for administrative subpoenas are not applied stringently.²¹ Furthermore, judges have avoided interpreting the language in Section 5562 to constrain the Bureau’s actions: For example, Section 5562(c)(1) seems to assume that the Bureau’s CIDs are seeking information relevant to a violation (which would logically require the Bureau to identify a violation), and calls for a description of the conduct at issue (which one might expect to be specific enough for the CID recipient to find useful).²² But courts often defer to the Bureau in statutory or constitutional challenges to CIDs, accepting boilerplate lists of laws and vague descriptions of conduct.²³ Some courts ask for more specificity, but not often enough to alter the Bureau’s practice.²⁴

Until the Supreme Court revisits its lax review of administrative subpoenas, CID reform must come from legislators.²⁵

Problems with the CFPB’s Exercise of CID Authority

Critics with wide experience of the CFPB’s CIDs (including former Bureau staff) show that they are often poorly justified and impose unreasonable costs on recipients.²⁶ Bureau CIDs tend to be broad and vague. A CID might refer to possible violations of a list of federal laws such as the Truth in Lending Act and “any other federal consumer financial law.”²⁷

Some CIDs have failed to describe the conduct under investigation at all, or describe it only in general terms.²⁸ Recipients may struggle for years to understand what conduct they ought to correct.²⁹ The Bureau, when made aware of these problems, rarely answers questions to clarify the demand’s scope.³⁰

Worse, one attorney notes that the Bureau issues CIDs to learn about a market or a business’s procedures even when no violation of law is suspected.³¹ Or the Bureau may use CIDs to review all of a business’s activity “because the target operates in a business that is disfavored by the Bureau (e.g., small dollar lending),” requiring the business “to prove it has not violated a myriad of consumer financial laws.”³² The Bureau’s 2021 policy manual stipulates that in its investigation it may “seek assurance that a violation has not occurred.”³³ A CID issued with this purpose would require the recipient to prove a negative—an impossible task.³⁴

Given the scope of Bureau CIDs, recipients must produce evidence on a daunting scale. The Bureau may ask about matters going back many years, and unrealistic deadlines have compounded the burden placed on the recipient.³⁵

Recipients of the CFPB’s CIDs expend significant resources, sometimes enough to put them out of business.³⁶ For a small business, compliance costs run from tens of thousands of dollars to over a hundred thousand dollars; larger firms may incur over a million dollars in costs.³⁷

Bureau policies also discourage CID recipients from petitioning for relief. The Bureau treats the issuance of a CID as confidential but publicizes petitions. Petitioners must risk harm to their reputation for nothing, as petitions are rarely granted.³⁸ Similar agencies, such as the Federal Trade Commission (FTC) and the Securities and Exchange Commission (SEC), also publish petitions, claiming that this promotes transparency of the regulatory process and allows other subpoena recipients to learn from the public record.³⁹ However, transparency could be preserved along with confidentiality by redacting proper names from published petitions.⁴⁰ Meanwhile, the threat of publicity makes the petition a weak accountability mechanism. A general confidentiality rule would stop the Bureau from disadvantaging CID recipients that wish to petition the Bureau.⁴¹

That the Bureau’s powers and practices are modeled after those of other agencies is no defense, because subpoena practices at other agencies are also problematic.⁴² Furthermore, several critics compare the Bureau’s practice unfavorably to that of the Department of Justice (DOJ), the FTC, the SEC, and other agencies.⁴³ The FTC staff itself has recommended changes to the Bureau’s practice, as has the Arizona state attorney general.⁴⁴

Some commentators dismiss critiques of the CFPB, noting that lawbreakers naturally object to being investigated and lawyers seek to complicate investigations to generate larger bills.⁴⁵ But some CID recipients have never been shown to have broken any law.⁴⁶ Critiques are more likely to be understated than overstated: Commentators widely acknowledge that regulated businesses often avoid publicly critiquing (or suing) their regulator out of fear of retaliation.⁴⁷ The detailed examples and common themes that persist across the many complaints against the Bureau show that testimony about problems with CIDs should be taken seriously.

Between 2017 and 2019, the CFPB amended its practice manuals in response to an audit, case law, and public comments.⁴⁸ For example, the recommendation that CIDs describe the supposed violation and conduct in “very broad terms” was formally replaced by a call for narrower CIDs.⁴⁹

However, 2025 testimony shows that problems highlighted by earlier commentators have persisted.⁵⁰ Legislators have made no changes to the Bureau’s statutory CID authority, and nothing stops future Bureau leaders from indulging in significant CID overreach.

Why Safeguarding Prosperity and Liberty Requires CID Reform

In a criminal trial, the accused benefits from a presumption of innocence. And in a civil trial, a plaintiff cannot proceed without some evidence.⁵¹ But our legal system has no presumption of innocence during pretrial investigations, because investigators may begin with little evidence. According to conventional wisdom, investigators should enjoy broad discretion because limitations on that discretion might impede legitimate inquiries.⁵²

This understanding of investigators’ needs arose before Americans became so beset by bureaucracy. After World War II, when the Supreme Court abandoned meaningful scrutiny of administrative subpoenas, the Code of Federal Regulations (CFR) numbered less than 10,000 pages.⁵³ As of 2025, the CFR was 61,584 pages (down from 107,262 in 2024).⁵⁴ The occasions on which someone might be required to produce massive amounts of information with little or no justification have vastly multiplied. Artificial intelligence (AI) could enable agencies to issue CIDs and similar subpoenas on an even greater scale.⁵⁵ Agencies might also tire of issuing individual CIDs and force regulated firms into a “consolidated audit trail,” which functions like a daily production order and establishes nearly limitless surveillance of regulated financial enterprises and their customers.⁵⁶

The scope of regulators’ inquiries has also expanded along with the scale: Increasingly, laws and regulations do not serve an essential role in supporting markets, as traditional laws that prevent concrete harms such as fraud do. Instead, regulators use their legal authority to interfere with markets, often trying to fine-tune outcomes. Such fine-tuning will rarely, if ever, benefit consumers; this issue aside, fine-tuning rules are not necessary.⁵⁷ The argument that investigators need broad investigative powers is attenuated when the investigation concerns the (perhaps remote) possibility that an inessential rule has been violated. Put differently, the idea that investigators must have discretion to make legitimate inquiries loses force when policymakers are careless about what amounts to a legitimate inquiry.

Even when the evidence collected using an administrative subpoena is not later used in a criminal proceeding, subpoena-enabled fishing expeditions destroy businesses and ruin people’s lives.⁵⁸ Administrative subpoenas can be used to target political opponents.⁵⁹ This frenetic regulatory activity is too often out of proportion to any plausible benefits. Over time, the growth of regulation—and carelessness about its substance—means a level of cumulative agency overreach that should not be tolerated in a constitutional republic.

Some commentators justify the post-1940s Supreme Court’s lax scrutiny of administrative subpoenas by equating administrative agencies with grand juries.⁶⁰ But the risk of abuse of authority by agencies is far greater. Grand juries are convened under the neutral authority of a judge, not an agency pursuing its own agenda. And the grand jury is focused on evidence related to obtaining an indictment, which agencies such as the CFPB may not be.⁶¹ Moreover, grand juries are populated by ordinary people with different incentives and ideas of fairness than regulators. Thus, a grand jury comprised of ordinary citizens is a safeguard of liberty.⁶² Administrative agencies cannot be expected to play this role.

Given the Bureau’s history of egregious CID overreach, policymakers should prioritize statutory reforms, starting by adopting limits known to be workable in other investigative contexts.

Current Proposals for CID Reform Would Bolster Accountability

Current reform proposals seek to curb the Bureau’s issuance of overbroad CIDs. Some propose authorizing the Bureau to issue demands only when it has a reasonable suspicion that a violation has occurred. The reasonable-suspicion standard is applied in police stop-and-frisk searches and some, though not all, subpoenas of personal information;⁶³ it is less stringent than the probable-cause standard that police must satisfy to obtain a warrant. Representatives Andy Barr (R‑KY), Vicente Gonzalez (D‑TX), and Jared Moskowitz (D‑FL) have introduced the Civil Investigative Demand Reform Act of 2025 (CIDRA), which would require CIDs to make “specific reference to particular facts.”⁶⁴ This would nudge the threshold upward toward the reasonable-suspicion standard.⁶⁵ A higher threshold would be good policy. If criminal investigators must show probable cause before issuing a warrant, the CFPB should be able to manage the less stringent requirement that it articulate “particular facts” to justify its issuance of a subpoena.⁶⁶

Some proposals seek to reduce the burden of large-scale demands. CIDRA would require that CIDs be issued no later than six years after the alleged violation.⁶⁷ Such time limits should be workable; law enforcement agencies and prosecutors routinely contend with temporal limits such as statutes of limitations.

Other reforms aim to heighten Bureau accountability by allowing CID recipients to appeal directly upon the Bureau’s denial of a petition for relief from a CID.⁶⁸ Currently, a CID recipient obtains judicial review by refusing to comply with the CID and waiting for the CFPB to sue to enforce it.

CIDRA links provisions addressing burden and review. The act would permit petitioners to assert that a CID is “unduly burdensome, disproportionately expensive, and outside the scope of the inquiry” or “unreasonably cumulative or duplicative or can be obtained from some other source that is more convenient, less burdensome, or less expensive.”⁶⁹ And CIDRA would make the Bureau’s denial of a petition directly subject to judicial review on the petitioner’s initiative. Combined, these provisions have the potential to counteract petitioners’ hesitancy to litigate by signaling to courts that legislators seek meaningful, less deferential limits on CIDs and better chances of success for petitioners.

More Extensive Reforms Would Be Desirable

Recent Supreme Court precedents have called for less judicial deference to agencies’ interpretations of the law and their application of law to facts.⁷⁰ However, some judicial deference to investigators’ framing of facts is likely to persist. To improve prospects for fundamental change, policymakers should consider the additional reforms set out below. These changes would include detailed statutory language and/or refer to familiar standards. The first two reforms could be implemented by amending the current language of CIDRA; the third and fourth would require additional, independent legislation.

First, any reform calling for review of a “burdensome” CID should define “burdensome” to ensure that judges do not set the bar to relief too high. Current constitutional cases require the recipient to show that the burden of compliance “threatens to unduly disrupt or seriously hinder normal operations of a business.”⁷¹ If judges were to apply this standard to determine the meaning of “burdensome” in a reform statute, CID recipients would face an unduly high bar to relief.

Second, any reform that addresses the proportionality of the burden imposed by a CID should state that the Bureau must prove proportionality. For example, CIDRA requires consideration of whether the burden of compliance is “disproportionately expensive.” This reform would be strengthened by language explicitly stating that the Bureau should bear the burden of proof in showing that the expense is in proportion to the severity of the violation alleged.⁷² This will encourage the Bureau to develop an understanding of the severity of the violation it is concerned with early on.

Third, reforms should curb the Bureau’s tendency to issue overwhelming demands by requiring the Bureau to show that the materials it seeks are meaningfully relevant. Currently, the standard of relevance that CIDs must satisfy is low.⁷³ Note that if a CID’s description of the violation under investigation is broad, almost any document might be relevant. Therefore, relevancy reform must first address overbreadth. The Bureau’s CIDs should be required to (1) detail the specific conduct at issue; (2) cite the specific violation of law that the Bureau believes the conduct violates; and (3) allow challenges to CIDs asking for information not relevant to that specific alleged violation.⁷⁴ “Relevance” should be aligned with civil discovery standards under Federal Rule of Civil Procedure 26.⁷⁵ This new relevancy standard would be more stringent than the relevancy standard for grand juries, an appropriate measure given that grand juries are less likely to indulge in overreach, as noted above.⁷⁶

Fourth, a neutral arbiter—ideally a decisionmaker outside the agency—should decide petitions for relief from CIDs.⁷⁷ The CFPB director or others involved in enforcement and rulemaking should not decide petitions.⁷⁸ An Article III judge or a body of external administrative law judges could do so instead. Short of that, administrative judges within the Bureau could review petitions provided that a robust set of structural measures were implemented to protect their independence.⁷⁹

Conclusion

The CFPB has indulged in significant CID overreach. Investigators’ nearly unbounded subpoena powers harm legitimate enterprises, with few proportional gains to consumers in many cases. The scale and scope of the harm has grown along with the scale and scope of regulation. Legislators and courts should view the assumption that investigators need broad discretion much more skeptically.

Reform should begin by incorporating standards familiar from Fourth Amendment cases, such as requiring the Bureau to articulate particular facts sufficient to support a reasonable suspicion. The use of detailed language to signal to judges that legislators seek meaningful review of burdensome CIDs would likely curtail judicial deference. Reforms that support the Bureau’s accountability to neutral arbiters such as independent administrative law judges should also be made. Successful reforms could inform changes to the administrative subpoena process at other agencies.

Ultimately, there is a limit to how effective any procedural reform of the CID process will be, because the rampant growth of regulation has given regulators a pretext for investigating almost any actor in the market—for almost any reason or no reason at all. Fundamental change will require policymakers to avoid unnecessary rules and focus on defining rights truly foundational to the free operation of markets.

Citation

Singleton, Solveig. “Addressing Civil Investigative Demand Overreach at the Consumer Financial Protection Bureau,” Policy Analysis no. 1018, Cato Institute, Washington, DC, May 19, 2026.

Solveig Singleton

Policy Analyst, Center for Monetary and Financial Alternatives, Cato Institute

Notes

1. James Martin et al., “Law Governing Civil Investigative Demands in the United States,” Law Library of Congress, 2018, p. 1; and United States v. Markwood, 48 F.3d 969, 976 (6th Cir. 1995).

2. See generally Requests for Information: Bureau Civil Investigative Demands and Associated Processes, 83 Fed. Reg. 3686 (January 26, 2018); and A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. (March 26, 2025) (statement of Rebecca E. Kuehn, Partner, Hudson Cook).

3. Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562.

4. Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562(c)(2).

5. Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562(c)(3)(A) (“describe each class of documentary material to be produced under the demand with such definiteness and certainty as to permit such material to be fairly identified”).

6. Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562(d)(1).

7. See Brent Skorup, “Will ‘Administrative Subpoenas’ Survive?,” Cato at Liberty (blog), Cato Institute, February 5, 2026.

8. Richard A. Nagareda, “Compulsion ‘to Be a Witness’ and the Resurrection of Boyd,” New York University Law Review 74 (1999): 1575, 1581, 1594.

9. US Const. amend. V.

10. See, for example, In re Seper, 705 F.2d 1499, 1501 (9th Cir. 1983); and United States v. Ward, 448 US 242 (1980).

11. Shapiro v. United States, 335 US 1 (1948); Braswell v. United States, 487 US 99 (1988); Boyd v. United States, 116 US 616 (1887); and Richard A. Nagareda, “Compulsion ‘to Be a Witness’ and the Resurrection of Boyd,” New York University Law Review 74 (1999): 1578 (which describes the Supreme Court’s current limits on Fifth Amendment protection for subpoenaed documents as a “considerable mistake”). See also Brief of the Cato Institute and Investor Choice Advocates Networks as Amici Curiae in Support of Petitioners at 12, American Securities Association v. Securities and Exchange Commission, 147 F.4th 1264 (11th Cir. 2025) (No. 23–13396) (stating that cases declining to protect corporate records are inconsistent with the common law at the time of the American Founding), citing Fisher v. United States, 425 US 391, 418 n.4 (1976) (Brennan, J., concurring) (“Without a doubt, the common-law privilege against self-incrimination in England extended to protection against the production of incriminating personal papers prior to the adoption of the United States Constitution”); and Samuel A. Alito Jr., “Documents and the Privilege Against Self-Incrimination,” University of Pittsburgh Law Review 48 (1986–1987): 65 (“English precedents at the time of the adoption of the fifth amendment extended the protection of the privilege against self-incrimination to corporate as well as individual records”).

12. See Doe v. United States, 487 US 201, 220 (1988); and United States v. Hubbell, 530 US 27 (2000). See also Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 446 (which describes “act of production cases,” in which the act of providing a document is testimonial because it reveals the existence of the document).

13. Richard A. Nagareda, “Compulsion ‘to Be a Witness’ and the Resurrection of Boyd,” New York University Law Review 74 (1999): 1578, 1595, 1625–26.

14. A business might get temporary relief from a CID if an individual corporate officer asserts the right against self-incrimination and there is a parallel criminal proceeding. See, for example, Consumer Financial Protection Bureau v. Global Financial Support, Inc., Case No. 15-cv-02440-GPC-WVG (S.D. Cal. May 17, 2016) (Order granting stay of CFPB’s civil proceeding pending resolution of criminal proceeding).

15. Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 446; Shapiro v. United States, 335 US 1, 17–18 (1948) (setting out the required records doctrine); and Baltimore City Department of Social Services v. Bouknight, 493 US 549, 555–60 (1990).

16. US Const. amend. IV.

17. Oklahoma Press Publishing Co. v. Walling, 327 US 186, 194–95, 202–05 (1946) (demand for books, records, and papers by subpoena was a “constructive” search). See also McLane Co. v. Equal Employment Opportunity Commission, 581 US 72, 84 (2017).

18. United States v. Morton Salt Co., 338 US 632, 652 (1950) (subpoena must be authorized by statute, reasonably relevant, and not too indefinite); United States ex rel. Accardi v. Shaughnessy, 347 US 260, 266–67 (1954); United States v. Powell, 379 US 48, 51 (1964); and Linde Thomson Langworthy Kohn & Van Dyke, P.C. v. Resolution Trust Corporation, 5 F.3d 1508, 1513 (D.C. Cir. 1993) (stating that the court’s “role is limited to evaluating whether the subpoena is for a lawful purpose, whether the documents requested are relevant to that purpose, and whether the demand is reasonable”).

19. See v. City of Seattle, 387 US 541, 544 (1967) (stating that “the Fourth Amendment requires that the subpoena be sufficiently limited in scope, relevant in purpose, and specific in directive so that compliance will not be unreasonably burdensome”); Federal Trade Commission v. Church & Dwight Co., 665 F.3d 1312, 1315–16 (D.C. Cir. 2011) (stating that courts defer to agency’s interpretation of its own authority and the agency’s determination of the relevance of subpoenaed material); and Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 441–42 (which notes that few cases find an agency imposed an undue burden).

20. United States v. Stuart, 489 US 353, 360–61 (1989) (explaining that agency must issue the subpoena in good faith and not for an “improper purpose such as harassment,” which would include the “acquisition of bargaining power in a collateral dispute”); Donovan v. Lone Steer, Inc., 464 US 408, 415 (1984); and City of Los Angeles v. Patel, 576 US 409, 420 (2015).

21. The Administrative Procedure Act does not further limit administrative subpoenas. See Administrative Procedure Act, 5 U.S.C. § 555(d) (stating that agency may use subpoenas if “authorized by law”).

22. Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562(c)(1), (c)(2) (stating that the Bureau may issue a CID “whenever the Bureau has reason to believe that any person may be in possession, custody, or control of any documentary material or tangible things, or may have any information, relevant to a violation” and that “each civil investigative demand shall state the nature of the conduct constituting the alleged violation which is under investigation and the provision of law applicable to such violation”).

23. See Consumer Financial Protection Bureau v. Heartland Campus Solutions ECSI, No. 18–1516, 2018 Westlaw 3831444, at *6 (3d Cir. August 13, 2018) (case designated nonprecedential); Consumer Financial Protection Bureau v. Seila Law LLC, 923 F.3d 680, 684–85 (9th Cir. 2019), vacated on other grounds; Seila Law LLC v. Consumer Financial Protection Bureau, 591 US 197 (2020), on remand; and Consumer Financial Protection Bureau v. Seila Law LLC, 984 F.3d 715, 720 (9th Cir. 2020) (affirming original appellate court’s ruling that CID was not overbroad).

24. See Consumer Financial Protection Bureau v. Accrediting Council for Independent Colleges & Schools, 854 F.3d 683, 689–92 (D.C. Cir. 2017); and Consumer Financial Protection Bureau v. Source for Public Data, 903 F.3d 456, 458–60 (5th Cir. 2018).

25. Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 421–78. See also Richard A. Nagareda, “Compulsion ‘to Be a Witness’ and the Resurrection of Boyd,” New York University Law Review 74 (1999): 1575–1659.

26. See, for example, From Watchdog to Attack Dog: Examining the CFPB’s Chopra-Era Assault on Disfavored Industries, Hearing Before the House Committee on Financial Services, Subcommittee on Oversight and Investigations, 119th Cong., 1st sess. 4 (June 26, 2025) (written testimony of James Kim, Partner, Cooley LLP). (“In the absence of proper mechanisms to challenge or limit the scope of CIDs, companies have been subjected to defective, unreasonable, and overly broad CIDs that were ripe for challenge.”)

27. A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. 8–9 (March 26, 2025) (statement of Rebecca E. Kuehn, Partner, Hudson Cook).

28. See Consumer Financial Protection Bureau v. Source for Public Data, 903 F.3d 456 (5th Cir. 2018); Joann Needleman, National Creditors Bar Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0063, Consumer Financial Protection Bureau, April 26, 2018, p. 8; and Jan Stieger, Receivables Management Association International, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0076, Consumer Financial Protection Bureau, April 26, 2018, pp. 3–4.

29. See, for example, Ronald S. Canter, the Law Offices of Ronald S. Canter, LLC, and John H. Bedard Jr., Bedard Law Group, PC, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0019, Consumer Financial Protection Bureau, February 13, 2018.

30. A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. 9 (March 26, 2025) (statement of Rebecca E. Kuehn, Partner, Hudson Cook).

31. A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. (March 26, 2025) (statement of Rebecca E. Kuehn, Partner, Hudson Cook).

32. A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. 7 (March 26, 2025) (statement of Rebecca E. Kuehn, Partner, Hudson Cook). See also From Watchdog to Attack Dog: Examining the CFPB’s Chopra-Era Assault on Disfavored Industries, Hearing Before the House Committee on Financial Services, Subcommittee on Oversight and Investigations, 119th Cong., 1st sess. 3–7 (June 26, 2025) (statement of Jennifer Bassett, CEO, Pacific Rim Alliance Corporation).

33. Policies and Procedures Manual, Office of Enforcement, Version 3.2 (Consumer Financial Protection Bureau, February 2021), p. 37.

34. See Dodd–Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. § 5562(c)(1) (stating that CID may be used to collect information “before the institution of any proceedings”); and United States v. Powell, 379 US 48, 57 (1964) (holding that agency may investigate merely upon suspicion).

35. A New Era for the CFPB: Balancing Power and Reprioritizing Consumer Protections, Hearing Before the House Committee on Financial Services, Subcommittee on Financial Institutions, 119th Cong., 1st sess. 10 (March 26, 2025) (testimony of Bryan A. Schneider, Partner, Manatt, Phelps & Phillips, LLP); and Tom Quaadman, Center for Capital Markets Competitiveness, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0070, Consumer Financial Protection Bureau, April 26, 2018, p. 14.

36. Eric Ellman, Consumer Data Industry Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0085, Consumer Financial Protection Bureau, April 26, 2018, p. 2.

37. Joann Needleman, National Creditors Bar Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0063, Consumer Financial Protection Bureau, April 26, 2018, p. 10; Joel L. Incorvaia, Incorvaia & Associates, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0017, Consumer Financial Protection Bureau, February 13, 2018, p. 2 (firm was never informed of what conduct was being investigated and incurred costs of $110,000 in employee time alone); and Eric Ellman, Consumer Data Industry Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0085, Consumer Financial Protection Bureau, April 26, 2018, p. 2.

38. Edward DeMarco, Housing Policy Council of the Financial Services Roundtable, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0077, Consumer Financial Protection Bureau, April 26, 2018, p. 10; and Virginia O’Neill, American Bankers Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0052, Consumer Financial Protection Bureau, April 24, 2018, p. 11 (petitions routinely denied).

39. Donald S. Clark, Federal Trade Commission, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0037, Consumer Financial Protection Bureau, March 27, 2018, pp. 13–14 (questioning credibility of transparency arguments, noting that the Bureau uses information strategically to undermine CID targets).

40. See Edward DeMarco, Housing Policy Council of the Financial Services Roundtable, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0077, Consumer Financial Protection Bureau, April 26, 2018, p. 6; and Christopher Willis, Ballard Spahr LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0088, Consumer Financial Protection Bureau, April 26, 2018, p. 9 n.2. Compare with Donald S. Clark, Federal Trade Commission, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0037, Consumer Financial Protection Bureau, March 26, 2018, p. 15 (redaction too burdensome for agency).

41. See From Watchdog to Attack Dog: Examining the CFPB’s Chopra-Era Assault on Disfavored Industries, Hearing Before the House Committee on Financial Services, Subcommittee on Oversight and Investigations, 119th Cong., 1st sess. 5 (June 26, 2025) (statement of Jennifer Bassett, CEO, Pacific Rim Alliance Corporation) (noting the Bureau redacted information about breadth of CID from petition but published sensitive information).

42. See, for example, Federal Trade Commission Act, 15 U.S.C. § 57b‑1.

43. Christopher Willis, Ballard Spahr LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0088, Consumer Financial Protection Bureau, April 26, 2018, p. 7 (comparing the Bureau to the DOJ and FTC); Jan Stieger, Receivables Management Association International, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0076, Consumer Financial Protection Bureau, April 26, 2018, pp. 2, 4 (comparing the Bureau to the FTC and state regulators); and Virginia O’Neill, American Bankers Association, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0052, Consumer Financial Protection Bureau, April 24, 2018, pp. 4, 6, 8, 10 (comparing Bureau practice to the FTC and SEC).

>44. See Donald S. Clark, Federal Trade Commission, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0037, Consumer Financial Protection Bureau, March 26, 2018, pp. 13–14; and Nick Dranias on Behalf of Mark Brnovich, Attorney General AZ, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0095, Consumer Financial Protection Bureau, January 26, 2018.

45. See Christopher Peterson, Consumer Federation of America, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0069, Consumer Financial Protection Bureau, March 27, 2018, pp. 7, 15.

46. See, for example, Joel L. Incorvaia, Incorvaia & Associates, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0017, Consumer Financial Protection Bureau, February 13, 2018.

47. James Chareq, Hudson Cook, LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0075, Consumer Financial Protection Bureau, April 26, 2018, pp. 1–2 (clients are “concerned that the Bureau, under different future leadership, will use the Bureau’s vast supervision and enforcement authority to retaliate against any CID target that responds to the RFI and describes its actual CID experiences under the Bureau’s prior leadership”); and Lucy C. Bartholomew, “Effective Legal Challenge to Agency Action: Litigating Against the CFPB in the Post-Chevron Environment,” North Carolina Banking Institute 29, no. 1 (2025): 15 (which notes that litigation against the CFPB could lead to retaliation).

48. See “CFPB Announces Policy Change Regarding Bureau Civil Investigative Demands,” press release, Consumer Financial Protection Bureau, April 23, 2019; Office of the Inspector General, “The CFPB Generally Complies with Requirements for Issuing Civil Investigative Demands but Can Improve Certain Guidance and Centralize Recordkeeping,” 2017-SR-C-015, Federal Reserve Board of Governors, September 20, 2017; and Consumer Financial Protection Bureau v. Accrediting Council for Independent Colleges & Schools, 854 F.3d 683 (D.C. Cir. 2017).

49. See Policies and Procedures Manual, Office of Enforcement, Version 3.0 (Consumer Financial Protection Bureau, May 2017), p. 68; and Policies and Procedures Manual, Office of Enforcement, Version 3.2 (Consumer Financial Protection Bureau, February 2021), p. 65.

50. See, for example, From Watchdog to Attack Dog: Examining the CFPB’s Chopra-Era Assault on Disfavored Industries, Hearing Before the House Committee on Financial Services, Subcommittee on Oversight and Investigations, 119th Cong., 1st sess. (June 26, 2025) (written testimony of James Kim, Partner, Cooley LLP).

51. See Bruce L. Hay, “Allocating the Burden of Proof,” Indiana Law Journal 72, no. 3 (1997): 652, 654 (which notes that the plaintiff is almost always required to satisfy the burden of proof). For example, to survive early efforts to dismiss a case such as a defendant’s motion for summary judgment, a plaintiff must produce enough evidence to show that there is a genuine issue of material fact. See also Anderson v. Liberty Lobby, Inc., 477 US 242 (1986); and “Preponderance of the Evidence,” Legal Information Institute, Cornell Law School, March 2022 (to win a case, the plaintiff must prove its claims by a preponderance of the evidence, meaning that the fact finder must be satisfied that it is more than 50 percent probable that the claims are true).

52. See Prentiss Cox et. al. (Legal Academics), Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0102, Consumer Financial Protection Bureau, April 25, 2018.

53. “Page Count by Category,” Federal Register, National Archives. Some page counts include those for 1940 (5,307); 1943 (17,553); 1947 (8,902); and 1949 (7,952).

54. “Page Count by Category,” Federal Register, National Archives.

55. Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 469.

56. See generally Brief of the Cato Institute and Investor Choice Advocates Networks as Amici Curiae in Support of Petitioners, American Securities Association v. Securities and Exchange Commission, 147 F.4th 1264 (11th Cir. 2025) (No. 23–13396).

57. Examples of rules that amount to attempts to fine-tune markets include price regulation, capital rules, and liquidity rules. Regulators may claim to be addressing a market failure, but government failure is more probable, and attempts to address the so-called failure often make matters worse. See Scott Lincicome, “When the Regulatory Cure Is Worse than the Market Disease,” The Dispatch, August 28, 2024; and Ryan Bourne, “How ‘Market Failure’ Arguments Lead to Misguided Policy,” Cato Institute Policy Analysis no. 863, January 22, 2019. Also see generally Gordon Tullock et al., Government Failure: A Primer in Public Choice (Cato Institute, 2002).

58. See Graham Hughes, “Administrative Subpoenas and the Grand Jury: Converging Streams of Criminal and Civil Compulsory Process,” Vanderbilt Law Review 47, no. 3 (1994): 573, 579 (“If civil penalties have become the preferred sanction in many regulatory cases, these penalties now play an important deterrent and perhaps punitive role that, in earlier times, was the exclusive province of the criminal law. Their invocation and their impact often may be hard to distinguish from the imposition of a criminal sanction”).

59. Brent Skorup, “Will ‘Administrative Subpoenas’ Survive?,” Cato at Liberty (blog), Cato Institute, February 5, 2026. See also, for example, Media Matters for America v. Federal Trade Commission, No. 1:25-cv-5302 (D.C. Cir. 2025) (upholding preliminary injunctions against the FTC’s use of sweeping CID to retaliate against media company for speech critical of Elon Musk and X).

60. Charles Doyle, “Administrative Subpoenas in Criminal Investigations: A Brief Legal Analysis,” RL33321, Congressional Research Service, December 19, 2012, pp. 8–9.

61. Costello v. United States, 350 US 359, 362 (1956).

62. Clark Neily, “The Year of the Jury,” Cato at Liberty (blog), Cato Institute, December 29, 2025. See also United States v. Williams, 504 US 36, 47 (1992) (describing the grand jury, explaining that it functions “as a kind of buffer or referee between the government and the people”).

63. Tom Quaadman, Center for Capital Markets Competitiveness, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0070, Consumer Financial Protection Bureau, April 26, 2018, p. 14. See also Terry v. Ohio, 392 US 1 (1968); In re McVane, 44 F.3d 1127 (2nd Cir. 1995); Resolution Trust Corporation v. Walde, 18 F.3d 943, 949 (D.C. Cir. 1994) (subpoena of personal information applying “articulable suspicion” standard); In re Sealed Case, 42 F.3d 1412, 1417 (D.C. Cir. 1994); Freese v. Federal Deposit Insurance Corporation, 837 F. Supp. 22 (D.N.H. 1993); and Parks v. FDIC, No. 94–2262, 1995 Westlaw 529629, at *8 (September 13, 1995), withdrawn from publication following grant of rehearing (holding that the FDIC must “articulate a reasonable suspicion of wrongdoing” before a subpoena for personal financial records can be enforced).

64. Civil Investigative Demand Reform Act, H.R. 1653, 119th Cong. (2025).

65. See, for example, Maryland v. Buie, 494 US 325, 332 (1990) (requiring suspicion based on “specific and articulable facts” to justify a search); and Navarette v. California, 572 US 393, 397 (2014) (articulable facts).

66. Aram A. Gavoor and Steven A. Platt, “Administrative Investigations,” Indiana Law Journal 97, no. 2 (2022): 471 (“More careful Fourth Amendment scrutiny would deter the agency from being overbroad or needlessly intrusive from the start”). See also Harriman v. Interstate Commerce Commission, 211 US 407, 419–20 (1908) (holding that administrative subpoenas may be issued only in “cases where the sacrifice of privacy is necessary—those where the investigations concern a specific breach of the law”); and Federal Trade Commission v. American Tobacco Company, 264 US 298, 305–7 (1924) (stating that “anyone who respects the spirit as well as the letter of the Fourth Amendment would be loath to believe that Congress intended to authorize one of its subordinate agencies to sweep all our traditions into the fire [citation omitted] and to direct fishing expeditions into private papers on the possibility that they may disclose evidence of crime.… The analogies of the law do not allow the party wanting evidence to call for all documents in order to see if they do not contain it. Some ground must be shown for supposing that the documents called for do contain it”).

67. Civil Investigative Demand Reform Act, H.R. 1653 § 2(a), 119th Cong. (2025).

68. See generally Darren M. Welch and Bryan A. Burcat, “Legislation Introduced to Overhaul CFPB Civil Investigative Demand Process,” Skadden, Arps, Slate, Meagher & Flom LLP, December 4, 2024. Hesitancy to petition for relief from CIDs could be addressed by amendments to the Equal Access to Justice Act, 28 U.S.C. § 2412(d), which requires agencies to pay a victorious petitioner’s costs unless “the position of the United States was substantially justified or that special circumstances make an award unjust.” The Supreme Court ruled that “substantially justified” means “having a reasonable basis in law and fact,” despite some evidence that legislators intended to hold the government to a higher standard. See Pierce v. Underwood, 487 US 552, 565–67 (1988). A higher standard could be specified by statutory amendment. The CFPB has rarely, if ever, paid such costs. Finally, see also “EAJA Awards,” Equal Access to Justice Act Reporting.

69. Civil Investigative Demand Reform Act, H.R. 1653 § 2(e)(C), 119th Cong. (2025).

70. See Loper Bright Enterprises v. Raimondo, 603 US 369 (2024); and Garland v. Cargill, 602 US 406 (2024). Compare with City of Arlington v. Federal Communications Commission, 569 US 290 (2013) (holding that deference was due to agency’s interpretation of its own authority).

71. Federal Trade Commission v. Texaco, Inc., 555 F.2d 862, 882 (D.C. Cir. 1977); and Commodity Futures Trading Commission v. Ekasala, 62 F. Supp. 3d 88, 94 (D.D.C. 2014) (“The standard for showing that a request is unduly burdensome … is a high one”).

72. Bearing the burden of proof would mean that the agency must show by a preponderance of the evidence that it is more likely than not that its assertion of proportionality is correct. “Burden of Proof,” Legal Information Institute, Cornell Law School, November 2024.

73. See Securities and Exchange Commission v. Arthur Young & Company, 584 F.2d 1018, 1028–29 (D.C. Cir. 1978) (stating that agency may request documents by subpoena that are not “plainly irrelevant”); Resolution Trust Corporation v. Walde, 18 F.3d 943, 946 (D.C. Cir. 1994) (stating that agency interpretation of relevance accepted if not “obviously wrong”); and Federal Trade Commission v. Invention Submission Corporation, 965 F.2d 1086, 1090 (D.C. Cir. 1992) (describing standard of relevancy as more relaxed in investigative context than in judicial context, and burden of showing irrelevance is on the subpoena target).

74. See Christopher Willis, Ballard Spahr LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0088, Consumer Financial Protection Bureau, April 26, 2018, p. 4.

75. See also Christopher Willis, Ballard Spahr LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0088, Consumer Financial Protection Bureau, April 26, 2018, p. 4, citing 42 U.S.C. § 2000e–8(a) (requiring EEOC administrative subpoena to be “relevant to the charge under investigation”).

76. See United States v. R. Enterprises, Inc., 498 US 292, 301 (1991) (in grand jury cases, there must be a reasonable possibility that material subpoenaed will produce information relevant to the general subject of the investigation).

77. Reforms to other elements of the CFPB’s structure have been proposed to address CID overreach. For example, several commentators have proposed that CIDs be more carefully vetted by senior leadership. See, for example, James Chareq, Hudson Cook, LLP, Comment on Request for Information Regarding Bureau Civil Investigative Demands and Associated Processes, CFPB-2018–0001-0075, Consumer Financial Protection Bureau, April 26, 2018, p. 3.

78. “Administrative Conference Recommendation 2016–4: Evidentiary Hearings Not Required by the Administrative Procedure Act,” Administrative Conference of the United States, December 13, 2016, p. 4.

79. See Christopher J. Walker, “A Reform Agenda for Administrative Adjudication,” Regulation 44, no. 1 (Spring 2021): 30–35 (which describes proposals to house administrative law judges in the legislative branch, in the executive branch but outside existing agencies, or in the judicial branch); and William Yeatman, “In Defense of Incrementalism: A Response to Prof. Greve’s Proposal for Administrative Courts,” Cato at Liberty (blog), Cato Institute, February 15, 2019 (which notes that Congress could insulate administrative law judges at agencies like the CFPB from direct review by the agency, and describes a proposal to create an independent administrative judiciary).

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.