Introduction
The Department of Justice (“DOJ” or the “Department”) has announced the Fraud Oversight through Careful Use of Statistics (“FOCUS”) Initiative to encourage proactive engagement with data miners seeking to use their tools to support qui tam claims under the False Claims Act (“FCA”). The initiative reflects DOJ’s effort to engage with—and manage—the growing number of FCA relators who base their claims on analysis of public information rather than insider knowledge. It is also emblematic of the growing relevance of artificial intelligence (“AI”) and machine learning to the FCA.
Key Takeaways
DOJ’s FOCUS initiative may strengthen partnerships with sophisticated, knowledgeable data miners and encourage additional cases. It is also possible that the initiative signals growing hostility to less sophisticated data miners who are merely speculating based on publicly available information.
As DOJ and relators consider evolving generative AI tools for offensive use in FCA cases, companies receiving federal funds should also consider incorporating AI tools in proactive compliance and risk management activities:
- Know Your Data. Companies should assume that potential relators are analyzing all public-facing information, such as loan or grant awards, procurement data, and billing data. Consider whether the company can leverage its AI tools to analyze public and non-public information to identify inconsistencies or errors before submitting claims or proposals or to evaluate risk after submission.
- Know the weaknesses in data mining claims. Given the growing prevalence of data mining qui tams, companies should be mindful of the inherent weaknesses in data-based theories when responding to investigations and lawsuits. In responding to investigations, identifying alternative explanations and evidence negating scienter may be particularly powerful in convincing DOJ to decline the case. Once the relator is confirmed to be a data miner, early engagement with DOJ about its views on the public disclosure bar and whether DOJ will consider exercising its dismissal authority may be helpful in identifying a path to a quick resolution.
Background
When Congress enacted the FCA during the Civil War, its goal was to provide a financial incentive for insiders to report fraud impacting the government. During World War II, DOJ became frustrated that many relators were basing their lawsuits on public information—in some cases, by copying criminal indictments drafted by DOJ. Since then, Congress has amended the FCA several times with the goal of achieving “the golden mean between adequate incentives for whistle-blowing insiders with genuinely valuable information and discouragement of opportunistic plaintiffs who have no significant information to contribute of their own.”[1] While qui tam lawsuits present a potential opportunity for the government to recover money, they also create a burden for the Department, which has to investigate each lawsuit, and for companies that have to incur costs to respond to investigations of meritless complaints.
Over the past three years, the number of FCA lawsuits filed by relators has increased significantly. The 980 qui tam lawsuits filed in 2024 represented a 30% increase over the prior record high, and 2025 saw a similar increase to 1,297 qui tam lawsuits.[2] This record is set to be broken again in 2026, with over 780 qui tam lawsuits filed in the first half of fiscal year 2026.[3] DOJ largely attributes this increase to data miners, who have filed more than 45% of all qui tam complaints since FY2024.[4]
What is the FOCUS Initiative?
Under this initiative, DOJ is offering an opportunity for data miners to meet and discuss their techniques with the Civil Fraud Section, which handles FCA enforcement. Department officials say the goal of this initiative is “to improve the Department’s ability to prioritize working with the most successful data miners” by understanding their capabilities and evaluating how their efforts reliably correlate with uncovering actionable fraud. [5]
DOJ outlined several considerations for data mining relators that are relevant to “the effective use of enforcement resources,” including whether a relator has:
- Established “high-quality, reliable, and predictive data analyses and signals and a thorough understanding of the relevant legal obligations,”
- Complied with Rule 9(b)’s heightened pleading standard, which requires relators “to state with particularity the circumstances constituting fraud,”
- Assessed “potential alternative explanations for the observed conduct and . . . articulate[d] how the data, in combination with other available evidence, suggests both scienter and falsity,” and
- Understood and articulated “program eligibility requirements and relevant regulatory frameworks.”[6]
Reading between the lines, it appears that DOJ may be trying to conduct due diligence to enable it to more quickly triage qui tam lawsuits brought by data miners. The announcement touts DOJ’s goal of partnering with the most successful data miners, but it also emphasizes that DOJ “will allocate its resources to the most promising avenues for combating fraud and recovering taxpayer dollars.”[7] Department officials are silent as to what that resource allocation will entail, but it has options. It can allocate fewer resources to data miners who do not meet its standards, potentially resulting in more cursory investigations and quick declinations in some subset of cases. DOJ also has discretion to allow defendants to pursue dismissal under the FCA’s public disclosure bar[8] or to dismiss cases outright.[9]
Data Mining in FCA Cases
Data-based FCA cases are not new, and the government also remains very interested in leveraging its own non-public data to identify potential cases. The Department of Health and Human Services Office of Inspector General (“HHS-OIG”) has long lauded the ability of its Data Analytics Team “to identify and target potential fraud schemes and areas of program waste and abuse.”[10] DOJ’s Opioid Fraud and Abuse Detection Unit applied data-mining techniques to uncover and prosecute opioid-related healthcare fraud, obtaining settlements that resolved FCA allegations in several matters, including the acceptance of illegal kickbacks.[11] And HHS-OIG recently concluded a research challenge to explore new AI applications to detect Medicare fraud.[12]
Private data miners have had some success in bringing healthcare claims based on publicly available Medicare data, and as noted by DOJ, have played a significant role in triggering FCA settlements arising out of the Paycheck Protection Program (“PPP”) and pandemic assistance loans.[13] PPP cases are particularly well-suited to data mining, as loan data is publicly available and many of the reasons for ineligibility (e.g., size, type of business, or Chinese government ownership) can also be gleaned from publicly available information.
But without any inside information, data miners face challenges in bringing viable FCA claims. In the context of data mining cases, multiple appellate courts have noted that “a statistical trend that is [also] consistent with a plausible alternative (and legal) explanation” does not constitute a viable FCA claim.[14] Consequently, DOJ appears to be pushing data mining relators to provide more substance than just a possible statistical theory.
Dinsmore’s multi-disciplinary FCA team is always available to discuss DOJ’s enforcement priorities and possible compliance strategies to identify and minimize risks.
[1] United States ex rel. Springfield Terminal Ry. v. Quinn, 14 F.3d 645, 649 (D.C. Cir. 1994)
[2] See Dep’t of Just., False Claims Act Settlements and Judgments Exceed $6.8B in Fiscal Year 2025 (Jan. 16, 2026), https://www.justice.gov/opa/pr/false-claims-act-settlements-and-judgments-exceed-68b-fiscal-year-2025.
[3] See Dep’t of Just., Focus Initiative for Data Miners Filing Qui Tam Complaints (Apr. 30, 2026), https://www.justice.gov/opa/media/1438871/dl.
[4] Id.
[5] Id.
[6] Id.
[7] Id.
[8] See 31 U.S.C. § 3730(e)(4) (requiring dismissal, unless the government objects, if substantially the same allegations or transactions as alleged in the action or claim were publicly disclosed in the news media, government report, or proceeding involving the government).
[9] See 31 U.S.C. § 3730(c)(2)(A) (providing the government may dismiss a qui tam action notwithstanding the objections of the relator after notice and opportunity for a hearing).
[10] See Jason Mehta & Jennifer A. Short, Big Data Makes Big Cases: How Data Analytics Is Shaping False Claims Act Enforcement, 67 The Fed. Law. (July/August 2020) 42, 44, https://www.fedbar.org/wp-content/uploads/2022/02/Pages-from-Mehta-Big-Data-Article.pdf.
[11] Id.
[12] Centers for Medicare & Medicaid Servs. (CMS), Crushing Fraud Chili Cook-Off Competition: Summary Report (Jan. 2026), https://www.cms.gov/files/document/crushing-fraud-chili-cook-white-paper.pdf.
[13] See Pandemic Response Accountability Comm., FRAUD PREVENTION ALERT: Using Data Analytics to Compare Income Representations by Applications Seeking Benefits from Multiple Federal Programs Could Have Prevented Hundreds of Millions of Dollars in Pandemic Fraud, pp. 1-2 (May 2025), https://www.pandemicoversight.gov/fraud-prevention-alert-income-misrepresentation; See also Dep’t of Just., supra note 2.
[14] Integra Med Analytics LLC v. Providence Health & Servs., 854 F. App’x 840, 844(9th. Cir. 2021); see also United States ex rel. Integra Med. Analytics, L.L.C. v. Baylor Scott & White Health, 816 F. App’x 892, 898 (5th Cir. 2020) (“[S]tatistical datacannot meet [the FCA’s] pleading requirements if, among other possible issues, it is also consistent with a legal and obvious alternative explanation.”).