[00:00:00] Speaker 03:
Our next case for argument is 25-1114.

[00:00:05] Speaker 03:
And, counsel, if I pronounce this wrong, you'll correct me. Mr. Oguntade versus MSPB.

[00:00:12] Speaker 01:
Of course, Your Honor. It's Mr. Oguntade.

[00:00:15] Speaker 03:
Say it again.

[00:00:16] Speaker 01:
Mr. Oguntade.

[00:00:17] Speaker 03:
Mr. Oguntade versus MSPB. All right. Mr. Harding, please proceed.

[00:00:23] Speaker 01:
Good morning, Your Honors, and may it please the Court, Alex Harding on behalf of Petitioner Babatunde Oguntade. Mr. Oguntade blew the whistle on an AI tool named CAPI that the VA had advertised as being designed and implemented to speed up veterans' claims processing by five days on average per claim. In reality, it was misclassifying so many claims that it was causing delays of on average five days and up to nine weeks in the case of complicated instances. Mr. Oguntade was found to not have jurisdiction for his whistleblower complaint below because the board found that his disclosures were not protected.

[00:00:58] Speaker 02:
Can I just ask you kind of more of a housekeeping question? Are their April and May disclosures materially different from the March disclosure?

[00:01:06] Speaker 01:
So, Your Honor, certainly the government concedes in footnote four of its response brief that it doesn't believe that there's any.

[00:01:12] Speaker 01:
With respect to whether or not those might have different connections to reprisal on remand, hasn't been decided.

[00:01:20] Speaker 02:
So, for example... What are you saying that the government concedes? You just mentioned something, and I think you... Kind of stop before telling me exactly what was being conceded.

[00:01:28] Speaker 01:
My apologies, Your Honor. So in footnote four of the government's brief, it notes that its arguments as to the March disclosure would apply equally to April and May and does not make separate arguments to them. We believe that the March disclosure is strong for its protected status. April and May would be even stronger, perhaps, because of the continued use of CAPI. But on remand, those may play out materially differently because the reprisal that Mr. Oguntade received from the VA may be different for each of those claims.

[00:01:59] Speaker 01:
So there's still material.

[00:02:00] Speaker 02:
But the government, I feel like, did concede that it was at least— preserved for the March disclosure, right?

[00:02:06] Speaker 01:
Correct, Your Honor.

[00:02:06] Speaker 02:
Okay. And that would be enough? Like, if we have one disclosure, we don't necessarily need to deal as much with April and May? Or what do you contend? You

[00:02:16] Speaker 01:
Of course, Your Honor. So this court could vacate and remand with instructions to adjudicate just the March disclosure or vacate and remand with instructions to adjudicate the April and May ones as well. But you do only need March in order to vacate. That's correct, Your Honor.

[00:02:32] Speaker 00:
The issue here is evidence of reasonable belief.

[00:02:39] Speaker 00:
objective with respect to waste or mismanagement. This was an experimental program. There was no assurance that it was going to work perfectly and not lead to delays. So why wasn't the board correct in dismissing this whistleblower complaint?

[00:02:59] Speaker 01:
Of course, Your Honor. So I'll tackle them one at a time if that's okay. So with respect to gross mismanagement, It's certainly the case that the government can try new programs and experiment with ways to improve its functionality. This isn't meant to be a referendum on that or even CAPI itself. It's only asking more narrowly whether when Mr. Oguntade made his disclosures, whether he was allowed to be fired or reprised on for them. So with respect to gross mismanagement, Mr. Oguntade, at the jurisdictional stage, the bar is very low, certainly could have a reasonable belief that this tool that's been deployed since 2019, so at the point he makes his March disclosures, it's already been working its way through hundreds of thousands of veterans' claims, at that point had caused delays of, as we were mentioning, up to nine weeks for disabled veterans.

[00:03:48] Speaker 01:
So this wasn't something that just launched. And I would also note that this isn't meant to be something that chills government trying new things. This is actually, with one small exception, an instance of the Whistleblower Protection Act working perfectly. Here was a government inefficiency that someone called out very specifically, offered evidence of the inefficiency, and it was resolved in July later that year when the tool was taken down. The only problem is The government fired him for it, or at least on his allegations, that's where we are for the jurisdictional stage.

[00:04:20] Speaker 01:
With respect to gross waste of funds, the standard there, I know there's different numbers for different cases in this court's precedent.

[00:04:28] Speaker 00:
Was he fired, or did he have a term appointment and didn't get reappointed?

[00:04:33] Speaker 01:
Right. So, Your Honor, it's correct that the two adverse actions being pled here are that he was non-renewed, and that he was terminated early, two separate ones. It's sort of a pretextual firing when we say it because the terms for presidential innovation fellows are renewed year by year. However, in the context of a whistleblower claim, this court has been clear that the scope of what counts as an adverse action is broader than normal. So I could give you a quote, Your Honor, from I believe it's from Johnston.

[00:05:04] Speaker 01:
that the board's jurisdiction does not normally extend to personnel actions such as reductions in job responsibilities. Under the WPA, however, a very broad class of personnel actions falls within the board's jurisdiction. Thus, changes in work duties and disciplinary action are personnel actions that are within the board's jurisdiction if they resulted from disclosures protected by the WPA, and that's page 912.

[00:05:25] Speaker 02:
Council, what case do you think is closest in terms of finding a relatively small harm to be significant for 2302 purposes.

[00:05:35] Speaker 01:
I think Hasami is our most analogous case for a few reasons. One, it covers both gross mismanagement and gross waste. There, the issue was a dispute in standard of care for hepatitis C patients, and there was an allegation by the whistleblower that too much money was being spent, and it was burning up the budget too quickly so that other drugs couldn't be delivered. But there, unlike here, at least there was good that was being done. Somebody still is receiving hepatitis C, treatment at the end of the day. Whereas here, at least at the pleading stage, we're not sure what benefit the government could point to that isn't purely speculative, with Mr. Oguntade showing that, again, the only thing that CAPI was supposed to be doing that was good was speeding up claims, but it was doing the exact opposite on them.

[00:06:21] Speaker 01:
So that bar, at least for gross waste of funds, measuring the amount of money spent against the benefit expected to accrue, we would be dividing by a very small benefit, if any.

[00:06:32] Speaker 02:
You're, of course, also point out there are these two options, gross waste of funds or gross mismanagement. Which do you think he has a stronger case under?

[00:06:40] Speaker 01:
Well, we certainly think that they are both. We think that they're both strong.

[00:06:45] Speaker 01:
I standing here, I don't know that I could pick between one or the other. I think gross mismanagement may be the most intuitive between the two, and I haven't answered your question with respect to that yet, so I will if I could. There, the VA's own mission statement, both on its website and on the job posting that Mr. Ogentade applied to, is to deliver claims faster. And this tool that they were using, and perhaps unknowingly at the time, was literally doing the opposite of that. So the gross mismanagement standard is, for a disclosure to be protected under the gross mismanagement standard, it has to show a substantial risk of significant adverse impact to the agency's mission.

[00:07:25] Speaker 01:
Many of the cases we cited are about risks, something that hasn't materialized yet. Here, though, for gross mismanagement, Mr. Ogentade showed that with statistical analysis complete with P values to a statistical certainty that this thing was doing the exact opposite. We're past the risk stage. It's already causing the harm. And so, you know, I know the government makes a point in its brief that this is only five days. The VA is already quite slow. It takes 100 days on average for veterans to get their benefits for their medical care.

[00:07:56] Speaker 01:
But I just want to reemphasize that, and we made this point in our brief, For the most complicated cases, for veterans with the most disabilities who need care the most, we're seeing delays up to nine weeks as a result of this AI tool misclassifying things, Your Honor.

[00:08:11] Speaker 03:
And the $1.5 million, right?

[00:08:13] Speaker 01:
And the $1.5 million, exactly, Your Honor. When the VA believed that it was saving five days per claim, it couldn't stop talking about it. It had product briefs that were touting the importance of saving five days, how it was going to save $1.5 million in costs that it had saved already. There's elsewhere. in the appendix where they cite other numbers. But certainly by the VA's own admission, this was going to be substantial to save five days.

[00:08:36] Speaker 03:
To be clear, the VA itself said if we can save five days in processing time, it will amount to approximately $1.5 million cost savings in labor, correct?

[00:08:45] Speaker 01:
Correct, Your Honor.

[00:08:46] Speaker 03:
And so, therefore, if your client is showing not only are you not saving five days, you're actually causing more process and delays longer than that, then there's reason – at least at this stage of the proceeding, to credit the VA's own claims about how much the labor costs were based on time.

[00:09:07] Speaker 01:
Absolutely, Your Honor, completely. And that's even before he's gotten to discovery to find other things. At the jurisdictional stage, you're absolutely right. The bar is extremely low.

[00:09:16] Speaker 03:
Okay, why don't you save the rest of your time for rebuttal. Let's hear from Ms. Schabacher. Ms. Schabacher.

[00:09:36] Speaker 04:
Good morning. May it please the court. My name is Deanna Schaubacher, representing MSPB.

[00:09:41] Speaker 04:
In this case, the board properly accepted the petitioner's factual allegations as true, but found that he failed to non-frivolously allege a reasonable belief that his disclosures regarding CAPI evidenced gross mismanagement or gross waste of funds. Both of these categories of wrongdoing require a significant, undebatable error by the agency.

[00:10:00] Speaker 04:
Gross mismanagement requires the agency's action create a substantial risk of significant adverse impact on the agency's mission. As this court has held, a lawful but problematic policy may amount to gross mismanagement only if reasonable minds could not debate the error in policy. Mere differences of opinion about the best approach to a management problem do not meet that standard.

[00:10:22] Speaker 04:
Here, the petitioner disclosed that CAPI was underperforming upon its initial implementation.

[00:10:28] Speaker 04:
but this did not involve any undebatable substantial risk of adverse impact on the agency's ability to carry out its mission. Rather, his disclosure evidenced at most a debatable agency decision and a policy disagreement about the agency's approach, including its decision to continue developing CAPI notwithstanding early indications.

[00:10:47] Speaker 02:
At this stage, though, he just needs to make a non-frivolous allegation, right? Like, it sounds like you might be getting a little beyond to more of the merits focus. What do you say to that?

[00:10:58] Speaker 04:
Well, a non-frivolous allegation would be accepting his factual allegations as true, whether he disclosed a reasonable belief that this disclosure evidenced gross mismanagement or gross waste of funds. Even at the non-frivolous stage, he must make, he must, accepting as true, allege that the disclosed matter met this high threshold of being gross. The WPA doesn't cover just mismanagement or just waste of funds. It requires gross mismanagement, gross waste of funds, and as this court has held many times, that must be an undebatable high threshold.

[00:11:34] Speaker 04:
So he can't just disclose that there are some problems upon initial implementation in a system, and he must disclose that it's gross mismanagement or gross waste of funds. So even at the non-fulfillment stage, he still must meet that definition of the type of disclosure that he's making.

[00:11:52] Speaker 03:
Well, he's alleged that you're pursuing this CAPI classification system, and the VA is touting it as being so beneficial that it's saving $1.5 million, when in fact his analysis shows that it's miscategorizing things with such a high error rate that it's resulting in a pretty significant delay in processing claims, not an improvement. And so why can't he just use that same, not only are you not saving $1.5 million, but if that's what you would save, you know, with a five-day improvement and this system is, I've got some evidence that shows it's causing delays.

[00:12:33] Speaker 03:
Why can't I use that now? I mean, $1.5 million might not sound like a lot to you, but it does sound like a lot to me. I mean, I think that could qualify at least at this stage of the process.

[00:12:44] Speaker 04:
Well, before the board, he never made any allegation about this $1.5 million. He didn't preserve that claim before the board. That was something that he pointed to for the first time on review. So the administrative judge didn't have the benefit of the claim that it... It is in the appendix, right? Page 357? They did submit this slideshow that says, $1.5 million in savings, but this was just hidden in the record. He never pointed it out to the administrative judge or preserved that argument.

[00:13:17] Speaker 03:
But even if he did... Wait, you say he didn't preserve the argument. He just has to make a non-frivolous allegation, correct?

[00:13:27] Speaker 04:
Of a reasonable belief, yeah.

[00:13:28] Speaker 03:
Right, so he made this allegation... And he submitted evidence. You're right. In the complaint itself, he didn't point to the $1.5 million number. But is that the standard? Or is it enough if he says there's this gross waste of funds? And then he submits evidence that kind of shows that. At this stage, that's not enough? You think he has to have the $1.5 million number in the complaint?

[00:13:50] Speaker 04:
Well, he needs to alert the board of what he wants to be adjudicated. And in this case, he didn't actually allege that it was gross waste of funds before the board, but the board, the administrative judge said to the extent it amounts to gross waste of funds, he has failed to point to any disproportionate expense. So this claim was really not preserved before the board.

[00:14:12] Speaker 00:
$1.5 million out of the VA budget doesn't sound gross to me.

[00:14:20] Speaker 04:
No, probably not.

[00:14:21] Speaker 00:
Sounds like pocket change for the VA.

[00:14:26] Speaker 04:
I would agree that it doesn't quite rise to a gross waste of funds in the terms of the VA.

[00:14:33] Speaker 03:
But even just out of curiosity, if the VA doesn't think it's a big deal, then why do they have a press release touting it where they put it in bold and say it amounts to a 24 times increase in their processing? Why do they issue a press release touting it exclusively, first and foremost, in bold. It's the only thing in bold in the document. So if the VA didn't think it was important, I'm just curious what would have prompted them.

[00:14:59] Speaker 04:
Well, I'm not saying that it's not important, excuse me, but it's not reaching the threshold of gross waste of funds. How do we figure out what the threshold of gross waste of funds is? It has to be a non-debatable expenditure that's out of proportion with the benefit reasonably expected to accrue to the government.

[00:15:18] Speaker 03:
But if the VA itself is touting how great it is that they just saved $1.5 million and we're at a non-frivolous pleading stage, you don't think that we can accept that as clearly establishing that the VA thinks $1.5 million is a value, sufficient value under these circumstances? No.

[00:15:38] Speaker 04:
It doesn't show that it's a gross waste of funds. The expected benefit here would have been enormous to partially automate 1.5 million veterans benefits claims going forward each year.

[00:15:56] Speaker 02:
You cannot... But I still think, again, you're doing a little bit of conflating what needs to be shown for jurisdictional purposes versus what needs to be shown on the merits. So...

[00:16:06] Speaker 02:
We are not deciding here, right? It's not before us whether or not he per se wins on the merits. But I don't understand why he wouldn't satisfy the non-frivolous allegation standard, not to mention the fact that he was also pro se below. So that at least means we should construe it more liberally than if he were represented by counsel below, too.

[00:16:30] Speaker 04:
He still has the burden to establish jurisdiction before the board by making a non-frivolous allegation of a reasonable belief. And this court has repeatedly explained that these types of wrongdoing require a high threshold. So we're accepting his factual allegations as true. We're accepting his analysis that this new system upon its implementation was actually delaying claims.

[00:16:53] Speaker 02:
Is there a high threshold for the jurisdictional stage or just a high threshold to win on the merits?

[00:17:00] Speaker 04:
At the jurisdictional stage, because we're talking about the definition of gross mismanagement and gross waste of funds. His allegations must be accepted as true, whether they assert a gross waste of funds or gross mismanagement that's plausible on its face. And these...

[00:17:17] Speaker 04:
These slight delays while they're implementing, developing a new technology product that has potential enormous benefit for years to come, these kind of expected bugs while they're innovating and trying to improve the system going forward, and the petitioner is actually there to work on and help fix these bugs in the system, that just doesn't reach the plausible assertion, even accepting all that is true, that there's a gross mismanagement or gross waste of funds here.

[00:17:44] Speaker 00:
What about the chilling factor? that if we find that there was a protected disclosure here, it chills the willingness of an agency to experiment.

[00:17:59] Speaker 04:
I think that's true, that there would be, why would you try to improve upon the manual processing of these claims if within the first implementation it's not meeting those goals right away and someone, you know, claims that this was gross mismanagement and gross waste of funds.

[00:18:19] Speaker 04:
So I do believe that you're right, that it would have a chilling effect on agency innovation.

[00:18:28] Speaker 03:
I guess, yeah. Just out of curiosity, why would it have a chilling effect? Just don't fire the dude. Like, that seems that complicated, right? Like, you don't have to not use Cappy. You don't have to listen to him. Just don't fire him. Well, that's fair.

[00:18:43] Speaker 03:
All right. That was well done.

[00:18:45] Speaker 04:
Completely disarmed me.

[00:18:49] Speaker 04:
So, yes, the board correctly found that there was simply not a non-foolish allegation that met the threshold of gross mismanagement or gross waste of funds here. And their kind of repackaging of the claims on review should not be accepted. We're looking at what was presented before the administrative judge. It was just these marked disclosures.

[00:19:10] Speaker 04:
with no information about what kind of expenditure might have been significantly out of proportion with the benefit sought to accrue to the agency.

[00:19:17] Speaker 02:
And you agree it's a non-finalist allegation that someone could reasonably believe that they were disclosing gross mismanagement or gross waste of funds, right?

[00:19:28] Speaker 04:
Yes, but the definition of gross mismanagement and gross waste of funds requires an undebatable definition.

[00:19:35] Speaker 04:
error for mismanagement or an undebatably out-of-proportion expenditure with the benefit reasonably expected to accrue. So a reasonable belief of an undebatable error or a reasonable belief of a significantly disproportionate expenditure. Okay, anything further? Unless there's no further questions, we ask that the Court affirm the Board's decision. Thank you.

[00:20:04] Speaker 03:
Mr. Harding, you have some rebuttal time.

[00:20:11] Speaker 01:
Good morning, Your Honors. I just wanted to address a few final points before we close. First, the court was absolutely correct in noting that at the board for pro se applicants, their submissions are construed in the most liberal light. And as we pointed out in our briefs, Mr. Oguntade did point to his disclosures and to the VA materials that he cited, including that product brief. With respect to your Honor's question, I think you mentioned that the $1.5 million doesn't sound gross.

[00:20:42] Speaker 01:
And I just want to reiterate that the standard is not necessarily the number in the abstract, but the number as compared to the benefit reasonably expected to accrue. Now, you didn't hear my colleague refer to any benefit that was expected to accrue, and I don't know that they've raised any. So at least with respect to gross waste of funds, that $1.5 million has to be weighted.

[00:21:00] Speaker 00:
Let me ask you about the failure to renew the appointment or the removal.

[00:21:08] Speaker 00:
These relate to retaliation, right?

[00:21:12] Speaker 01:
Correct, Ron.

[00:21:12] Speaker 00:
Has that been argued, whether the retaliation was appropriate? Aren't we just talking about gross mismanagement?

[00:21:26] Speaker 01:
Absolutely, Your Honor. You do not need to decide whether he was retaliated against or any of that. You only need to make a decision as to the Board's finding on the protected status of his disclosures. Absolutely.

[00:21:38] Speaker 01:
One last point before I wrap, Your Honor. My colleague referred to not debatable several times in her argument. I believe that comes from this Court's decision in White 6. That's the 2004 decision. That decision was a gross mismanagement decision, and it was on the merits. I'm first not aware of any instance of that being used for gross waste of funds or at the jurisdictional stage. And as Your Honor has pointed out, the I can't emphasize this enough. Mr. Oguntade hasn't even gotten his day in court yet. He's only at the complaint stage. This court analogized the jurisdictional stage to the Iqbal Twombly standard in its Hasami decision, and that's all that he has to meet.

[00:22:15] Speaker 01:
All of the decisions that we've seen in this case so far where the whistleblower fails to meet the jurisdictional stage are because they lack specificity. That's cases like Young, cases like Kavanaugh here. It's hard to imagine how Mr. Okintade could have been more specific in the things that he disclosed.

[00:22:33] Speaker 01:
Finally, I just want to wrap by saying that CAPI is certainly one of the first AI tools that the government has tried to use to automate its processes. It's not a bad thing that it's trying to do that, but there will probably be many more. And in those instances where we want the tools to succeed, But they aren't. It's important that people be able to tell their supervisors that so that those can be fixed in a timely manner. And that's exactly what we believe the Whistleblower Protection Act was designed to protect. If your honors have no other questions.

[00:23:03] Speaker 03:
I thank both counsel. The case is taken under submission.