Why Humans Distrust Algorithms – and How That Can Change - Knowledge at Wharton

Recently, as I scrolled through my Google News Alert email for “False Claims Act,” one government press release caught my eye.  The title – “National Roofing Company Settles PPP Fraud Allegations for $9 Million” – while certainly noteworthy, was not what interested me.  Instead, it was the identity of the whistleblower:  Sidesolve, LLC, which the press release stated would receive a $1 million whistleblower bounty after settling a False Claims Act (FCA) case.  Sidesolve describes itself as an AI/data analytics company that uses “data science to find large scale corporate fraud in healthcare, such as kickbacks and fake medical equipment.”

The Sidesolve FCA case was brought by a well-known and respected FCA whistleblower law firm, which included the following in their press release announcing the historic settlement:

“An unusual twist to this case is that our relator, Sidesolve LLC, has no preexisting relationship with the defendant.

In other words, Sidesolve – which itself is nothing more than a Certificate of Organization somewhere on the Secretary of State’s web portal – did nothing more than program an algorithm to comb through publicly-available PPP data and uncover an anomaly that could, potentially, indicate fraud.  And, using the whole “even a broken clock is right twice a day” philosophy, if one were inclined to file enough of these cases against enough defendants, some of them are bound to stick, or at least result in a substantial settlement.

To be clear, I know these plaintiff attorneys personally, and can confirm that they are great people and outstanding lawyers, and I sincerely congratulate them on their continued success.  That being said, and with all respect to my friends, the Sidesolve settlement raises some serious questions regarding the purpose of the FCA’s whistleblower provisions.  Those provisions, also known as the qui tam provisions, permit whistleblowers to receive between 15 and 30% of the government’s eventual, often sizable, recovery. Do these “data Relators” actually serve the intended purpose of those provisions?  To put it bluntly, what unique and valuable role do these data Relators serve that justifies their receipt of six- and sometimes seven-figure whistleblower awards?  If the purpose of the FCA’s whistleblower provisions is to encourage corporate insiders to come forward with knowledge of fraud or wrongdoing, does an LLC with “no preexisting relationship with the defendant” deserve $1 million?

Before you assume that the Sidesolve case was “one of a kind,” without going into details for reasons related to the FCA’s seal provisions, I can personally attest to the fact that this is not the only data Relator case making its way through the FCA investigative process.  In fact, recently, I spoke on an FCA conference panel with another notable whistleblower attorney who touted his firm’s success in mining publicly available Medicare data and then, similar to Sidesolve, filing whistleblower complaints under the FCA on behalf of faceless LLCs.

To be fair, I would not go so far as to say these data Relators serve no purpose.  There is probably some value in having private individuals search through publicly available data and point out suspicious patterns for the government.  But, since the government has equal (and probably better) access to the exact same data, and also has plenty of very sophisticated computer scientists to help implement advanced algorithms in order to search that data without the need for a whistleblower, are these private data Relators truly earning their oftentimes substantial bounty? In a world where AI now allows anyone with a smartphone to create a sophisticated algorithm on voice command, I think that is a serious question that will need to be answered soon.

In addition to these policy questions, as technology advances and, inevitably, more data Relators emerge, courts will also be faced with whether such cases pass muster under the FCA’s “Public Disclosure Bar,” which requires dismissal of an FCA whistleblower action if the underlying allegations or transactions were publicly disclosed.  Given that federal courts generally interpret the public disclosure bar broadly, data Realtors will likely face a difficult time pursuing a declined FCA case where the complaint was based solely upon their combing of publicly available data.

To further illustrate this point, it is helpful to look at the Supreme Court’s decision in Schindler Elevator (131 S.Ct. 1885, 2011).  There, in holding that responses to FOIA requests were covered by the FCA’s public disclosure bar, the Court, in a decision authored by Justice Scalia, held that the case “seems to us a classic example of the ‘opportunistic’ litigation that the public disclosure bar is designed to discourage, because anyone could have filed the same FOIA requests and then filed the same suit.”  Could the Court say the same thing about a data Relator LLC, set up for the sole purpose of combing publicly available data in search of a statistical anomaly and, hopefully, a hefty whistleblower award?  For now, I’ll leave it with the old adage, “only time will tell.”

If you need representation in connection with a False Claims Act matter, contact us today.