United States Court of Appeals
for the Federal Circuit
______________________
PALANTIR USG, INC.,
Plaintiff-Appellee
v.
UNITED STATES,
Defendant-Appellant
______________________
2017-1465
______________________
Appeal from the United States Court of Federal
Claims in No. 1:16-cv-00784-MBH, Judge Marian Blank
Horn.
______________________
Decided: September 7, 2018
SEALED OPINION ISSUED: September 7, 2018
PUBLIC OPINION ISSUED: September 13, 2018 *
______________________
THEODORE OLSON, Gibson, Dunn & Crutcher LLP,
Washington, DC, argued for plaintiff-appellee. Also
represented by KAREN LOUISE MANOS, AMIR C. TAYRANI;
JOSH KREVITT, New York, NY; HAMISH HUME, STACEY K.
GRIGSBY, JON KNIGHT, JOSHUA RILEY, Boies, Schiller &
* This opinion was originally filed under seal and
has been unsealed in full.
2 PALANTIR USG, INC. v. UNITED STATES
Flexner, LLP, Washington, DC; DAVID BOIES, Armonk,
NY.
DOMENIQUE GRACE KIRCHNER, Commercial Litigation
Branch, Civil Division, United States Department of
Justice, Washington, DC, argued for defendant-appellant.
Also represented by CHAD A. READLER, ROBERT EDWARD
KIRSCHMAN, JR., DOUGLAS K. MICKLE.
GIDEON A. SCHOR, Wilson, Sonsini, Goodrich & Rosati,
PC, New York, NY, for amicus curiae Technology Net-
work. Also represented by ADAM WILLIAM BURROWBRIDGE,
Washington, DC.
______________________
Before NEWMAN, MAYER, and STOLL, Circuit Judges.
STOLL, Circuit Judge.
The government appeals from a permanent injunction
on its solicitation of bids for Distributed Common Ground
System – Army Increment 2 (“DCGS-A2”), the Army’s
primary system for processing and disseminating multi-
sensor intelligence and weather information. The United
States Court of Federal Claims granted the injunction
after concluding that the Army failed to comply with the
requirements of 10 U.S.C. § 2377. We affirm.
PALANTIR USG, INC. v. UNITED STATES 3
BACKGROUND 1
Palantir USG, Inc. (“Palantir”) filed a pre-award bid
protest in the Court of Federal Claims, challenging the
Army’s solicitation 2 for DCGS-A2. The solicitation seeks
a single contractor to be the system data architect, devel-
oper, and integrator of DCGS-A2. Palantir’s complaint
alleges that the Army violated § 2377(c) by, among other
things, failing to determine whether its needs could be
met by commercial items before issuing the contested
solicitation. See § 2377(c)(2). To provide background, we
introduce the applicable statute and regulations, the
DCGS-A2 system, the relevant facts regarding pre- and
post-solicitation activity, and the procedural history of
this case.
I. The Statutory and Regulatory Preference for the
Acquisition of Commercial Items
This appeal is centered on the Federal Acquisition
Streamlining Act (“FASA”), which requires that federal
agencies, to the maximum extent practicable, procure
commercially available technology to meet their needs.
Federal Acquisition Streamlining Act of 1994, Pub. L.
No. 103-355, § 8104, 108 Stat. 3243 (1994) (codified as
amended at 10 U.S.C. § 2377). As the government
1 The facts and procedural history of this case are
extensive. A more exhaustive recitation of the facts
underlying this appeal may be found in the Court of
Federal Claims’ opinion. See Palantir USG, Inc. v. United
States, 129 Fed. Cl. 218, 221–43 (2016) (“CFC Op.”). We
provide a summary of the relevant facts and procedural
history necessary to resolve this appeal.
2 United States Department of the Army, Army
Contracting Command, Aberdeen Proving Group’s Re-
quest for Proposals (“RFP”) No. W56KGY-16-R-0001.
4 PALANTIR USG, INC. v. UNITED STATES
acknowledges, the legislative history reflects Congress’s
understanding that “[t]he purchase of proven products
such as commercial and nondevelopmental items can
eliminate the need for research and development, mini-
mize acquisition leadtime, and reduce the need for de-
tailed design specifications or expensive product testing.”
S. Rep. No. 103-258, at 5 (1994), reprinted in
1994 U.S.C.C.A.N. 2561, 2566.
FASA provides that the “Federal Acquisition Regula-
tion [FAR] shall provide regulations to implement” FASA.
41 U.S.C. § 3307(e)(1). Under FASA:
(a) Preference.—The head of an agency shall en-
sure that, to the maximum extent practicable—
(1) requirements of the agency with respect to
a procurement of supplies or services are stat-
ed in terms of—
(A) functions to be performed;
(B) performance required; or
(C) essential physical characteristics;
(2) such requirements are defined so that
commercial items or, to the extent that com-
mercial items suitable to meet the agency’s
needs are not available, nondevelopmental
items other than commercial items, may be
procured to fulfill such requirements; and
(3) offerors of commercial items and nondevel-
opmental items other than commercial items
are provided an opportunity to compete in any
procurement to fill such requirements.
10 U.S.C. § 2377(a); see FAR 11.002(a)(2). FASA further
requires that agencies shall, “to the maximum extent
practicable . . . acquire commercial items or nondevelop-
PALANTIR USG, INC. v. UNITED STATES 5
mental items [NDIs] other than commercial items to meet
the needs of the agency.” § 2377(b)(1); see FAR 12.101(b).
FASA achieves its preference for commercial items in
part through preliminary market research. Before solicit-
ing bids or proposals, agency officials must conduct mar-
ket research 3 concerning the availability of commercial
items pursuant to § 2377(c)(1), which states:
(1) The head of an agency shall conduct market
research appropriate to the circumstances—
(A) before developing new specifications for a
procurement by that agency;
(B) before soliciting bids or proposals for a
contract in excess of the simplified acquisition
threshold; and
(C) before awarding a task order or delivery
order in excess of the simplified acquisition
threshold.
§ 2377(c)(1); FAR 10.001(a)(2). Next, agency officials
must use that market research to determine whether
commercial items can meet the agency’s requirements,
with or without modification of either the commercial
items or the agency’s requirements, pursuant to
§ 2377(c)(2):
3 The implementing regulations define market re-
search as “collecting and analyzing information about
capabilities within the market to satisfy agency needs.”
FAR 2.101(b). Conducting market research “involves
obtaining information specific to the item being acquired”
and the regulation explains that the “extent of market
research will vary, depending on such factors as urgency,
estimated dollar value, complexity, and past experience.”
FAR 10.002(b)(1).
6 PALANTIR USG, INC. v. UNITED STATES
(2) The head of an agency shall use the results of
market research to determine whether there are
commercial items or, to the extent that commer-
cial items suitable to meet the agency’s needs are
not available, nondevelopmental items other than
commercial items available that—
(A) meet the agency’s requirements;
(B) could be modified to meet the agency’s re-
quirements; or
(C) could meet the agency’s requirements if
those requirements were modified to a rea-
sonable extent.
§ 2377(c)(2) (emphasis added); FAR 10.001(a)(3). Palantir
argues—as it did below—that the Army violated both of
these FASA mandates, § 2377(c)(1) and (2).
II. The Distributed Common Ground System –
Army Increment 2 (DCGS-A2)
We briefly introduce the purpose and evolution of the
Army system at issue here. The Distributed Common
Ground System (“DCGS”) is made up of Army, Air Force,
Navy, and Marine Corps ground processing systems that
can share information across the Joint Force. The over-
arching purpose of the Army’s Distributed Common
Ground System (“DCGS-A”) is to combine all of the Ar-
my’s intelligence software/hardware capabilities into one
program with the ability to access and be accessed by
Army intelligence and command components, as well as
other military and intelligence systems. DCGS-A in-
cludes many software products—commercial, government,
and open source—as well as software integration that
allows all the different products and components to com-
municate and operate seamlessly. CFC Op., 129 Fed. Cl.
at 223.
PALANTIR USG, INC. v. UNITED STATES 7
The original DCGS-A Increment 1 (“DCGS-A1”) is op-
erational and deployed worldwide, but its “data architec-
ture is over 10 years old and is based upon technology
that is nearing obsolescence, with no growth margin.” Id.
at 233. Therefore, in 2014, the Army began investigating
the best way to approach DCGS-A2, which would “intro-
duce a new and modernized data management architec-
ture (DMA) using a modular system approach to perform
Army intelligence analysis capabilities.” Id. at 223.
The performance work statement (“PWS”) for this so-
licitation stated that the requirements of DCGS-A2 in-
cluded the ‘‘development of new data architecture,
standards based enhanced visualization and analytical
tools, cloud computing and ‘big data’ analytic capabilities;
cyber analytics and data integration, visualization capa-
bilities, Cyber Operations, Interoperability, Counter
Intelligence/HUMINT, Weather, GEOINT, Geospatial
Engineering and Sensor Management,’’ and explained
that ‘‘[t]hese efforts include Software Development, Capa-
bility Enhancements, Integration, Limited Fielding and
Training support, Maintenance, and Support for logistics
development, for a period of performance of six years from
contract award.’’ Id. The draft version of the perfor-
mance work statement for the DCGS-A2 solicitation
stated that ‘‘[t]he DCGS-A Increment approach utilizes
spiral deliveries to maintain interoperability with Army
and Joint ISR [Intelligence, Surveillance and Reconnais-
sance] architectures and to address capability insertion
and enhancements. This system must remain interoper-
able and compatible with the Joint command system
infrastructure and mission applications.’’ Id. As indicat-
ed by the contracting officer who issued the solicitation,
the data management architecture “will serve as the
architecture foundation and the heart with which the rest
of the capabilities will depend on to function. The [data
management architecture] development is therefore the
8 PALANTIR USG, INC. v. UNITED STATES
focus of the first task order executed under the DCGS-A
Increment 2 contract.’’ Id.
III. The Palantir Gotham Platform
As noted above, Palantir argues that the Army violat-
ed § 2377(c) by failing to determine whether the Army’s
needs could be met by commercial items, and that, had
the Army done so, it would have issued one or more
solicitations to procure commercial or nondevelopmental
items to satisfy the DCGS-A2 requirements. J.A. 194–95.
Specifically, Palantir submits that its flagship software
product, the Gotham Platform, could satisfy the Army’s
requirements.
Palantir’s Gotham software product is a data-
management platform that Palantir began to market to
private sector and government customers in 2009. The
software enables agencies to integrate, visualize, and
analyze large amounts of data from different sources that
reside in different databases in different formats. The
parties stipulate that the government has previously
procured the Palantir Gotham Platform on a commercial
item basis, and that the GSA Schedule lists both term
licenses and perpetual licenses for Palantir’s platform.
Parties’ Joint Stipulations of Fact ¶¶ 3–4, Palantir USG,
Inc. v. United States, No. 1:16-cv-00784-MBH (Fed. Cl.
Apr. 24, 2017), ECF No. 125 (“Joint Stip.”). The parties
also stipulate that Palantir Gotham is a commercially
available data management platform.
IV. Pre-Solicitation Activity
The parties do not dispute the following Court of Fed-
eral Claims’ fact findings. In 2014, the Army decided to
curtail the third release of DCGS-A1 and redirect its
efforts to acquiring and launching DCGS-A2. The Army
PALANTIR USG, INC. v. UNITED STATES 9
chartered an independent Data Integration, Visualization
and Analytics (“DIVA”) Market Study. 4 This market
study, dated July 2014, was completed by the MITRE
Corporation, a not-for-profit research and development
organization. According to the parties, the DIVA Market
Study was intended to “provide situational awareness and
market trends to the Army leadership of the ‘state-of-the-
practice’ within the commercial DIVA software platform
landscape.” Joint Stip. ¶ 11. The DIVA Market Study
report summarized MITRE’s recommendations for the
DCGS-A2 acquisition effort. According to the report, the
DIVA Market Study assessed three acquisition approach-
es:
a. Cloud Infrastructure Platform Provider: Pro-
vide highly-scalable and reliable computing infra-
structure services (e.g., data bases [sic]; analytic
engines; computing and storage; identity man-
agement);
b. Turn-Key: Procure a commercial product as ba-
sis of [DCGS-A2] infrastructure. Integrate addi-
tional applications onto this infrastructure[;]
c. Hybrid approach: both an Enterprise Cloud
Platform and a Turn-Key Platform, including in-
tegration of additional applications . . . .
Id. ¶ 11 (citations omitted). Having considered each of
these three acquisition approaches, the DIVA Market
Study recommended the hybrid approach. The study’s
“Key Observation” was that the hybrid approach blends
the benefits of the other two approaches, provides “the
global scale of the cloud infrastructure with the ‘out-of-
4 A DIVA platform is sometimes referred to as a
“data management platform.” Compare J.A. 18400, with
J.A. 12226–27.
10 PALANTIR USG, INC. v. UNITED STATES
the-box’ capabilities of the DIVA ‘Turn Key’ platform,”
and provides “better tactical edge support.” J.A. 12234.
The DIVA study outlined how to apply a hybrid approach
to the DCGS-A2 capabilities. In particular, the hybrid
approach would start with procurement of two Commer-
cial-off-the-Shelf (“COTS”) foundation components:
(a) cloud infrastructure services and (b) a DIVA “Turn
Key” infrastructure platform. Though the DIVA Market
Study did not analyze any potential vendors, it assessed
overall market trends and served as an early indication
that commercial items should be considered for the
DCGS-A2 infrastructure platform.
Following completion of the DIVA Market Study, the
Army issued three requests for information (“RFIs”). It
issued its RFI #1 in August 2014, just one month after
release of the DIVA Market Study. The goal of RFI #1
was to assess “the level of relevant competition and
capabilities in the market place and elicit industry feed-
back to assist the Program Office in developing the Acqui-
sition Plan” for the potential DCGS-A2 procurement.
J.A. 11802. It requested “respondents’ corporate overview
information and basic qualifications in managing software
development projects that are similar in scope and process
to the DCGS-A program.” J.A. 11876 (emphasis added);
see also J.A. 11876–81.
The Army issued RFI #2 in December 2014. Palantir
responded, expressing concern that RFI #2 was focused on
collecting information on the respondents’ ability to
conduct a large-scale development effort, instead of as-
sessing existing software capabilities that would be
applicable to DCGS-A1 capability gaps.
The Army issued RFI #3 in May 2015, which was
meant to “[i]nform the small business role for Increment 2
[and] [d]etermine if [a Small Business Set-Aside] is ap-
propriate.” J.A. 11803. In response, Palantir again
highlighted its concerns with the Army’s acquisition
PALANTIR USG, INC. v. UNITED STATES 11
approach, asserting that “[t]he successful delivery of
Increment 2 depends on the answer to a central question:
will the Army acquire a data platform from the commer-
cial market or will it attempt to build one itself?”
J.A. 11918.
In July 2015, the Army Materiel Systems Analysis
Activity issued a Trade Space Analysis, which identified
and evaluated technical functionality, cost, usability,
schedule risk, and technical risk for DCGS-A2. The
report indicated that the Trade Space Analysis would
inform the economic analysis and RFP for DCGS-A2 and
analyzed the following options: COTS, Government-off-
the-Shelf (“GOTS”), and hybrid. The report concluded
that a hybrid COTS-development approach was the best
of the three alternatives, noting that such an approach
was currently functioning in the Department of Defense
Intelligence Community and would only require minor
development to fill capability gaps.
On July 13, 2015, however, the Army issued a Market
Research Report that concluded the opposite—that “the
[DCGS-A2] development effort cannot be procured as a
commercial product.” J.A. 11840. This Market Research
Report indicated that three features were not available as
commercial products: Data Fusion, Intelligence Support
to Cyber, and DCGS Integrated Backbone Upgrade. Id.
It further addressed Palantir directly, finding Palantir’s
response to earlier RFPs non-responsive because Palantir
“did not provide any examples of past experience relevant
to the development of Increment 2.” J.A. 11835–36. The
Market Research Report further stated that, “[b]ased on
the Market Research to date, the recommended approach”
for DCGS-A2 “is a five (5) year Engineering and Manufac-
turing Development (EMD) effort consisting of two releas-
es.” J.A. 11841 (emphasis added). It also stated, without
any explanation, analysis, or support, that “[s]ignificant
portions of the anticipated Increment 2 scope of work” are
“not available as a commercial product.” J.A. 11840.
12 PALANTIR USG, INC. v. UNITED STATES
Two days later, on July 15, 2015, the Army issued its
draft performance work statement, defining the efforts
required to acquire services for the development and
integration of DCGS-A2. J.A. 10410–594. The perfor-
mance work statement defined the requirements for
DCGS-A2 to include “development of new data architec-
ture” and completion of the “design, development, integra-
tion and test.” J.A. 10418–19. Palantir filed a response to
the draft performance work statement in October 2015,
asserting that the “Army does not need to build that [data
management] platform, as it can buy it today.”
J.A. 10693. In the same month, on October 21, 2015,
Ms. Heidi Shyu, as the Senior Procurement Executive,
signed a Determination & Findings for “Award of a Single
Source Indefinite-Delivery Indefinite-Quantity (IDIQ)
Single Award Contract Exceeding $103M” for DCGS-A2.
J.A. 12298–304. The Determination & Findings noted
that DCGS-A2 “is heavily focused on design and develop-
ment of a new data management architecture by a con-
tractor as the systems integrator,” and “[d]evelopment of
the data integration layer is pivotal and complicated by
multiple interfaces and interoperability requirements
with external intelligence systems.” J.A. 12299 ¶ 4. The
Determination & Findings concluded that:
[I]ssuing a single award IDIQ contract will miti-
gate many of the risks identified herein and is in
the best interest of the Government. Due to the
complex developmental efforts this work entails,
further competition at the task order level would
interrupt development, ultimately increase price,
and cause schedule slippages.
....
[A] single-source task or delivery order contract
estimated to exceed $103 million for [DCGS-A2]
Engineering Manufacturing and Development
contract is authorized because the task or delivery
orders expected under the contract are so integral-
PALANTIR USG, INC. v. UNITED STATES 13
ly related that only a single source can reasonably
perform the work.
J.A. 12302, 12304.
V. The Solicitation
On December 23, 2015, the Army issued the solicita-
tion that is the subject of this appeal. The solicitation
contemplated the award of a single indefinite-delivery,
indefinite-quantity contract for DCGS-A2, with the simul-
taneous issuance of a cost-reimbursement type task order.
It sought a single contractor to be the system data archi-
tect, developer, and integrator of DCGS-A2. The solicita-
tion also required a software capability demonstration,
which the Army contemplated “could include a Govern-
ment Furnished Information (GFI), Commercial Off-[t]he-
Shelf (COTS), Government Off-the-Shelf (GOTS), or Open
Source product(s).” J.A. 10960. The performance work
statement accompanying the solicitation explained that
the successful offeror would be responsible for, among
other things, the development of new data architecture;
cloud computing and big data analytic capabilities; data
integration; and interoperability with counter intelli-
gence/human intelligence. The performance work state-
ment also stated that the software design
release/development should include “maximization of
reuse of GOTS/COTS products.” J.A. 11101 ¶ 3.4.1.
VI. Post-Solicitation Activity
Shortly after the Army issued the solicitation, Palan-
tir filed a pre-award bid protest, which the Government
Accountability Office (“GAO”) denied in May 2016. See
generally Palantir USG, Inc., No. B-412746, 2016 WL
3035029 (Comp. Gen. May 18, 2016) (“GAO Op.”). Then,
on June 30, 2016, Palantir filed the current pre-award bid
protest in the Court of Federal Claims. Count one of
Palantir’s complaint alleged that the Army violated
§ 2377 and 48 C.F.R. §§ 10.002 and 11.002 by refusing to
14 PALANTIR USG, INC. v. UNITED STATES
solicit the data management platform as a commercial
item. Count two alleged that the Army violated § 2377
and 48 C.F.R. §§ 10.002 and 11.002 by refusing to solicit a
commercial item for the entirety of DCGS-A2. Count
three alleged that the Army violated § 2377(c) by failing
to determine whether its needs could be met by commer-
cial items. On July 1, 2016, one day after Palantir filed
its complaint in the Court of Federal Claims, the govern-
ment issued a Determination of Non-Commercial Item,
laying out the government’s justification for its determi-
nation regarding commercial items following market
research.
In the Court of Federal Claims, the parties filed cross-
motions for judgment on the administrative record. The
Court of Federal Claims granted judgment in Palantir’s
favor, concluding that the Army failed to determine
whether commercial items meet or could be modified to
meet the agency’s needs and that, by failing to do so, the
Army acted in an arbitrary and capricious manner in
violation of 10 U.S.C. § 2377. CFC Op., 129 Fed. Cl.
at 282, 290. The court explained:
Not only did the agency fail to explain or indicate
what commercial items possibly were available or
had been considered, the Market Research Report
is devoid of any information regarding the possi-
ble commercial items that could be modified to
meet the Army’s requirements. . . . [T]here is no
evidence that the agency made [a determination
regarding the suitability of Palantir’s data man-
agement platform] after the market research was
complete or prior to issuing the solicitation. The
total absence of any discussion regarding commer-
cial items, or possible modifications to commercial
items, reinforces the court’s understanding that
the Army was focused on a developmental ap-
proach to the DCGS-A Increment 2 at an early
PALANTIR USG, INC. v. UNITED STATES 15
point in the procurement process, to the exclusion
of commercially available alternatives.
Id. at 276. Further finding that the Army’s actions
caused Palantir to suffer a ‘‘non-trivial competitive injury
which can be addressed by judicial relief,” the Court of
Federal Claims permanently enjoined the Army from
issuing a contract award under the protested solicitation.
Id. at 289–95 (quoting Weeks Marine, Inc. v. United
States, 575 F.3d 1352, 1362 (Fed. Cir. 2009)). According
to the injunction, the Army would have to properly and
sincerely comply with FASA § 2377 before awarding a
contract to meet its DCGS-A2 requirements. Id. at 295.
The United States appeals. We have jurisdiction under
28 U.S.C. § 1295(a)(3).
DISCUSSION
I. Standard of Review
We review the Court of Federal Claims’ ruling on the
parties’ cross-motions for judgment on the administrative
record de novo, applying the same standard of review as
the trial court. Glenn Def. Marine (Asia), PTE Ltd. v.
United States, 720 F.3d 901, 907 (Fed. Cir. 2013). Cross-
motions for judgment on the administrative record are
governed by Rule 52.1(c) of the Rules of the United States
Court of Federal Claims (“RCFC”). See RCFC 52.1(c). In
deciding these motions, the court considers “whether,
given all the disputed and undisputed facts, a party has
met its burden of proof based on the evidence in the
record.” A & D Fire Prot., Inc. v. United States,
72 Fed. Cl. 126, 131 (2006) (citing Bannum, Inc. v. United
States, 404 F.3d 1346, 1356 (Fed. Cir. 2005)).
The Army’s procurement decision must be set aside
only if it is “arbitrary, capricious, an abuse of discretion,
or otherwise not in accordance with law,” or “without
observance of procedure required by law.” 5 U.S.C.
16 PALANTIR USG, INC. v. UNITED STATES
§ 706(2)(A), (D). As the United States Supreme Court has
explained about § 706(2)(A) review:
[T]he court must consider whether the decision
was based on a consideration of the relevant fac-
tors and whether there has been a clear error of
judgment. Although this inquiry into the facts is
to be searching and careful, the ultimate standard
of review is a narrow one. The court is not em-
powered to substitute its judgment for that of the
agency.
Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402,
416 (1971), abrogated on other grounds by Califano v.
Sanders, 430 U.S. 99, 105 (1977) (internal citations omit-
ted). “Effective contracting demands broad discretion.
Accordingly, agencies ‘are entrusted with a good deal of
discretion in determining which bid is the most advanta-
geous to the Government.’” Lockheed Missiles & Space
Co. v. Bentsen, 4 F.3d 955, 958–59 (Fed. Cir. 1993) (quot-
ing Tidewater Mgmt. Servs., Inc. v. United States,
573 F.2d 65, 73 (Ct. Cl. 1978) (internal citations omitted)).
The reviewing “court’s task is to determine whether
‘(1) the procurement official’s decision lacked a rational
basis; or (2) the procurement procedure involved a viola-
tion of regulation or procedure.’” Savantage Fin. Servs.,
Inc. v. United States, 595 F.3d 1282, 1285–86 (Fed. Cir.
2010) (quoting Weeks Marine, 575 F.3d at 1358).
The government raises two issues on appeal:
(1) whether the trial court went beyond the statutory and
regulatory language of FASA and its implementing regu-
lations and imposed heightened obligations; and
(2) whether the trial court wrongly discarded the pre-
sumption of regularity and substituted its judgment in
determining that the Army acted arbitrarily and capri-
ciously and in violation of 10 U.S.C. § 2377. We address
each issue in turn.
PALANTIR USG, INC. v. UNITED STATES 17
II. The Trial Court Properly Concluded that the Army’s
Actions Violated FASA § 2377(c)(2)
The government first argues that the trial court erro-
neously added requirements to § 2377, including that the
Army was required to “fully investigate,” “fully explore,”
“examine,” and “evaluate” whether all or part of its re-
quirements could be satisfied by commercially available
items, such as Palantir’s product. Appellant Br. 35–36;
see also CFC Op., 129 Fed. Cl. at 282. We are not per-
suaded that the Court of Federal Claims imposed addi-
tional requirements beyond those required by the statute.
FASA requires an agency to use the results of market
research to “determine” whether there are commercial
items that “meet the agency’s requirements; could be
modified to meet the agency’s requirements; or could meet
the agency’s requirements if those requirements were
modified to a reasonable extent.” § 2377(c)(2). While the
trial court’s thorough opinion sometimes uses words other
than “determine,” we conclude that, read in context, those
words were intended to be synonymous with “determine.”
In any event, we need not devote significant discussion to
this argument, as we “sit to review judgments, not opin-
ions,” Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530,
1540 (Fed. Cir. 1983), and our de novo review leads us to
the same conclusion as the one reached by the Court of
Federal Claims.
As discussed above, we give deference to the Army’s
procurement decisions. Advanced Data Concepts, Inc. v.
United States, 216 F.3d 1054, 1058 (Fed. Cir. 2000). Even
with that deference, however, we conclude that the Ar-
my’s procurement actions in this case were arbitrary and
capricious and in violation of § 2377. First, as explained
in detail below, the administrative record demonstrates
that the Army, while conducting its market research, was
on notice of the desirability of hybrid options that used
commercial solutions and that Palantir claimed to have a
commercial item that could meet or be modified to meet
18 PALANTIR USG, INC. v. UNITED STATES
the Army’s needs. Furthermore, the record shows that
the Army did not use the results of that market research
to determine whether there were commercial items that
could meet its requirements, could be modified to meet its
requirements, or could meet its requirements if those
requirements were modified to a reasonable extent. See
§ 2377(c)(2).
The administrative record reflects that the Army was
on notice of the possibility that commercial items 5 could
satisfy its needs for portions of DCGS-A2. Indeed, the
Army was on notice as early as July 2014, when it re-
ceived the DIVA Market Study it had commissioned. As
discussed above, the DIVA Market Study recommended a
Phased Acquisition and Integration Approach as a poten-
tial strategy. In this phased strategy, the Army would
5 The term “commercial item” is defined at
41 U.S.C. § 103 and FAR 2.101, which apply government-
wide. 10 U.S.C. § 2302. Regarding products, a “commer-
cial item” is defined as an “item . . . that is of a type
customarily used by the general public or by non-
governmental entities for purposes other than govern-
mental purposes; and—(i) [h]as been sold, leased or
licensed to the general public; or (ii) [h]as been offered for
sale, lease, or license to the general public.” FAR 2.101.
The definition is broad enough to include products that
would meet the above provisions, but for “[m]odifications
of a type customarily available in the commercial market-
place,” or “[m]inor modifications of a type not customarily
available in the commercial marketplace made to meet
Federal Government requirements.” Id. “Minor modifica-
tions means modifications that do not significantly alter
the nongovernmental function or essential physical char-
acteristics of an item or component, or change the purpose
of a process.” Id.
PALANTIR USG, INC. v. UNITED STATES 19
first acquire the two foundation components: a COTS
cloud infrastructure service and a COTS DIVA “Turn
Key” platform. Next, integrating these two components
with each other and the DCGS-A Enterprise data man-
agement architecture would establish a baseline DCGS-
A2—a core suite of applications and analytics functions; a
new data management architecture. Indeed, the DIVA
Market Study explained that “[a] key advantage of lever-
aging COTS cloud infrastructure services and a COTS
DIVA platform is that doing so provides a significant
amount of technical infrastructure and end-user capabili-
ties.” J.A. 12251. Thus, the Army was aware of a possible
commercial approach for at least portions of the DCGS-A2
procurement.
Similarly, just six months before the solicitation, the
Army Materiel Systems Analysis Activity’s July 2015
Trade Space Analysis indicated that a hybrid approach
using commercial items was the best of three alternatives
considered, including COTS, GOTS, and hybrid. The
Trade Space Analysis described this hybrid approach as a
“compilation of commercially available software packages
augmented with integrated tools/widgets written by a
third-party using requirements/specifications generated
by the Government (i.e., combination of COTS and
GOTS).” J.A. 11954. Notably, the analysis acknowledged
that “[h]ybrid software option alternatives are currently
functioning in the [Department of Defense Intelligence
Community]” and would “only require minor development
to fill capability gaps.” J.A. 11976.
Palantir also put the Army on notice of its capabilities
to provide a commercial item that could be modi-
fied/integrated to meet the Army’s needs for DCGS-A2.
Palantir responded to the Army’s RFI #1 by explaining
that the Army should consider existing commercial solu-
tions:
20 PALANTIR USG, INC. v. UNITED STATES
The acquisition cycle should fully leverage exist-
ing commercial solutions. Prioritizing the rapid
procurement of commercial capabilities minimizes
the anticipated scope of development needed to
deliver Increment 2 capabilities. Narrowing the
development scope requires expanding the use of
commercially available COTS capabilities—it does
not require narrowing the overall scope of the
DCGS-A program. The Government does not need
to build Increment 2 functionality; the Govern-
ment can buy the core functionality from the
commercial market and integrate any number of
additional applications.
J.A. 11885 (footnote omitted). Palantir explained that
“we recommend the Government pursue a different acqui-
sition strategy than the strategy behind the Increment 1
challenges.” Id. Palantir further informed the Army that
it had successful contracts with the U.S. Marine Corps,
U.S. Immigration and Customs Enforcement, and the
Defense Intelligence Agency with its COTS solution.
Suggesting that the COTS approach would also work for
DCGS-A2, Palantir proposed using a firm-fixed-price
(“FFP”) model with an “outcomes-based Performance
Work Statement based on a proven product and incorpo-
rating support services.” J.A. 11889.
The Army’s December 2014 RFI Response Analysis
includes a summary of Palantir’s response:
Palantir has developed an intelligence fusion sys-
tem that has been used by various entities within
the Department of Defense. Palantir was found
capable to provide Data management and Work-
flow Management upgrades, and partially capable
of providing Data Fusion and Cyber capabilities to
Increment 2.
J.A. 11868. This confirms that the Army was aware of
Palantir’s commercially available intelligence fusion
PALANTIR USG, INC. v. UNITED STATES 21
system, which was already in use within the Department
of Defense, and considered Palantir capable of delivering
some of the required functionality of DCGS-A2.
Palantir again tried to explain the value of a commer-
cial—rather than developmental—approach in response
to the Army’s RFI #2:
We continue to believe that the success of Incre-
ment 2 requires a proven commercial solution to
ensure the delivery of a working capability on
time and within budget. We are concerned that
the present RFI . . . is focused on collecting infor-
mation on each respondent’s ability to conduct a
services-based, large-scale, and custom software
engineering effort . . . rather than to assess exist-
ing software capabilities applicable to Increment 1
capability gaps.
J.A. 11910. Likewise, Palantir’s response to RFI #3
explained that “[i]n cooperation with the government,
Palantir fields and manages 25 Palantir deployments at
every major Marine Corps command, representing over
15,000 accounts at peak usage across the Marine Intelli-
gence community.” J.A. 11922. Palantir further stated
that “Increment 2 should use a fielded commercial solu-
tion” and that “[d]elivering Increment 2 on a commercial
platform ensures the data layer advances at the same
pace as commercial technology.” J.A. 11918. Additional-
ly, Palantir flagged that it thought the Army’s “initial
decision to embark on a significant software development
effort, rather than acquiring a COTS solution,” would
cause challenges like the ones that faced DCGS-A1. Id.
In addition, the Administrative Record includes three
Operational Needs Statements from other Department of
Defense personnel requesting Palantir’s data manage-
ment platform. One such statement, dated February
2015, explained that “[t]he Palantir Command platform is
a proven capability that is currently in use to provide
22 PALANTIR USG, INC. v. UNITED STATES
COP, data integration, and staff integration capabilities
across multiple commercial and government organiza-
tions.” CFC Op., 129 Fed. Cl. at 224. It further stated
that Palantir “offers a solution that meets all of our
requirements.” Id.
Based on this record, we agree with the trial court
that the Army was, or should have been, aware of Palan-
tir’s data management platform. Despite repeated notice
that commercial products might well be available and
could be modified to meet the Army’s needs, the Army
concluded that DCGS-A2 could not be procured as a
commercial product with scant explanation. Indeed, the
Army’s July 2015 Market Research Report simply stated
that “[s]ignificant portions of the anticipated Increment 2
scope of work such as Data Fusion, Intelligence Support
to Cyber, and [DCGS Integrated Backbone] upgrade are
not available as a commercial product. As such, the
[DCGS-A2] development effort cannot be procured as a
commercial product.” J.A. 11840 ¶ 8.3.5; CFC Op.,
129 Fed. Cl. at 231. It concluded that “[b]ased on the
Market Research to date, the recommended approach is a
five (5) year Engineering and Manufacturing Develop-
ment (EMD) effort consisting of two releases.” J.A. 11841.
There was no discussion in the Market Research Report to
support the Army’s conclusory assessment that these
three requirements—data fusion, intelligence support,
and DCGS integrated backbone—were not commercially
available. Nor was there any discussion of whether any
commercial items could have been modified to meet the
Army’s needs or the Army’s requirements could have been
modified so that commercial items could be used.
Further, on the first requirement—data fusion—
record evidence shows that Palantir Gotham may provide
“data fusion” capability. The Army itself described Palan-
tir Gotham as being used as an “integrated fusion and
analysis platform.” J.A. 18183–84. In addition, Palantir
explained in its response to the Army’s draft performance
PALANTIR USG, INC. v. UNITED STATES 23
work statement that the commercial market offers nu-
merous existing tools with this capability. Palantir took a
step further and explained that this requirement, as
written, envisioned building the capability from scratch
instead of evaluating whether such functionality was
commercially available.
The record evidence likewise demonstrates that the
second requirement—intelligence support to cyber—may
have been commercially available. Indeed, the July 2013
MITRE Palantir Platform Information Brief, which is in
the administrative record, noted that “Palantir has
NETOPS capability to audit/log potential cyber events
and has Cyber Analysis Tools to detect/analyze suspicious
Cyber events.” J.A. 17851. Furthermore, Palantir ex-
plained in its response to the draft performance work
statement that having a separate requirement for cyber
intelligence functionality is unnecessary because the
Army could acquire such intelligence support to cyber
capabilities by simply acquiring the Palantir Gotham
Data Management Platform.
Finally, the record demonstrates that Palantir Go-
tham could be interoperable with the existing DCGS
integrated backbone. For example, Palantir contracted
with a U.S. military command to provide Gotham as an
information bridging solution, including to satisfy the
requirement that the data structure would support evolv-
ing DCGS integrated backbone standards and upgrades to
new versions. Furthermore, Palantir explained in its
response to the draft performance work statement that
the proposed requirement to integrate the DCGS inte-
grated backbone is unnecessary because it should be
treated as an interoperability standard, not a software
platform that should be integrated.
On this record, we agree with the trial court that the
Army failed in its obligation under § 2377 to determine
whether a commercial item could meet or be modified to
24 PALANTIR USG, INC. v. UNITED STATES
meet the Army’s procurement requirements. We
acknowledge that there is no statutory or regulatory
requirement for agencies to document their determina-
tions pertaining to § 2377 and FAR Part 10. See Ad-
vanced Am. Constr., Inc. v. United States, 111 Fed. Cl.
205, 227 (2013) (“[T]he language of section 10.002(e) is
precatory in nature and does not establish any mandatory
documentation requirement. That section states that
agencies ‘should’ document the results of their market
research; it does not state that those agencies ‘shall’ do
so.”). Nevertheless, the record must be sufficient to
permit meaningful judicial review consistent with the
Administrative Procedure Act, 5 U.S.C. § 706. See Motor
Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto.
Ins. Co., 463 U.S. 29, 43 (1983) (“[T]he agency must
examine the relevant data and articulate a satisfactory
explanation for its action including a ‘rational connection
between the facts found and the choice made.’” (quoting
Burlington Truck Lines v. United States, 371 U.S. 156
(1962))); see also Sierra Club v. Gorsuch, 715 F.2d 653,
660–61 (D.C. Cir. 1983) (“If there is reasoned deci-
sionmaking lurking behind such agency behavior, it is yet
to be articulated. For agency action to be upheld, it must
not only be explainable; it must also be explained.”);
Bagdonas v. Dep’t of Treasury, 93 F.3d 422, 426 (7th Cir.
1996) (“The statement of reasons need not include de-
tailed findings of fact but must inform the court and the
petitioner of the grounds of decision and the essential
facts upon which the administrative decision was based.”
(citing Kitchens v. Dep’t of Treasury, 535 F.2d 1197,
1199–1200 (9th Cir. 1976))); Impresa Construzioni Geom.
Domenico Garufi v. United States, 238 F.3d 1324, 1338
(Fed. Cir. 2001) (“Garufi”) (citing Supreme Court deci-
sions establishing that, even if the agency is not obligated
to provide reasons, a court may nonetheless order the
agency to provide explanation if such an explanation is
required for meaningful judicial review). Here, the ad-
ministrative record plainly shows that the Army was on
PALANTIR USG, INC. v. UNITED STATES 25
notice that Palantir’s product might be a commercial item
that would satisfy its requirements, whether as-is or with
modifications. Despite that notice, the Army’s ultimate
determination regarding its market research excluded
commercial items from consideration in a conclusory
fashion. On this record, we conclude that the Army did
not rationally use its market research results to deter-
mine whether there are available commercial items that:
“(A) meet the agency’s requirements; (B) could be modi-
fied to meet the agency’s requirements; or (C) could meet
the agency’s requirements if those requirements were
modified to a reasonable extent.” § 2377(c)(2);
FAR 10.001(a)(3)(ii).
The government argues that the Trade Space Analy-
sis demonstrates that it satisfied its obligations under
§ 2377. In particular, the government asserts that this
document shows that a hybrid approach—using “commer-
cially available software” and software developed by the
government (GOTS)—was superior to a commercial item
procurement. Appellant Reply Br. at 9. But the record
undermines the government’s position, showing that the
Army’s procurement efforts were focused on a develop-
mental approach without determining the viability of a
commercial or even a hybrid approach. Indeed, in its
Determination & Findings for Award of a Single Source
IDIQ Single Award Contract, the government emphasized
that DCGS-A2 “is heavily focused on design and develop-
ment of a new data management architecture by a con-
tractor as the systems integrator.” J.A. 12299 ¶ 4
(emphasis added). The government further emphasized
that “[d]evelopment of the data integration layer is pivot-
al.” Id. Nowhere in that document does the government
address implementation of the Trade Space Analysis’s
recommended hybrid approach or Palantir’s proposed
COTS approach. Moreover, although the Trade Space
Analysis recommended a hybrid approach using commer-
cial items, the government did not use that information to
26 PALANTIR USG, INC. v. UNITED STATES
determine whether a commercial item could be modified
to meet the agency’s requirements. See § 2377(c)(2). As
such, the Trade Space Analysis does not alter our conclu-
sion that the Army did not comply with § 2377.
III. The Trial Court Properly Accounted
for the Presumption of Regularity
We now turn to the government’s second ground for
challenging the trial court’s judgment. The government
alleges that the trial court wrongly discarded the pre-
sumption of regularity in determining that the Army’s
action was arbitrary and capricious and did not comply
with § 2377(c)(2). We do not agree.
Under the Administrative Procedure Act, even where
an explanation or reason is not required for an agency’s
determination, a reviewing court has the power to require
an explanation. Garufi, 238 F.3d at 1338. “[I]n determin-
ing whether to require an explanation, the agency deci-
sion is entitled to a presumption of regularity.” Id. (citing
Bowen v. Am. Hosp. Ass’n, 476 U.S. 610, 626–27 (1986)).
“Because of that presumption of regularity, the agency
should not be required to provide an explanation unless
that presumption has been rebutted by record evidence
suggesting that the agency decision is arbitrary and
capricious.” Id.
Here, the court extensively cited record evidence
showing that the Army’s decision was arbitrary and
capricious and in violation of 10 U.S.C. § 2377. In partic-
ular, the court performed a searching review and analysis
of the DIVA Study, Trade Space Analysis, RFIs and RFI
responses, July 2015 Market Research Report, Octo-
ber 21, 2015 Determination & Findings for Award of a
Single Source IDIQ Single Award Contract, and July 1,
2016 Determination of Non-Commercial Item. Based on
this review, it concluded that the Army neglected to
determine whether possible commercially available alter-
natives meet or could be modified to meet the require-
PALANTIR USG, INC. v. UNITED STATES 27
ments of the Army’s acquisition. See CFC Op.,
129 Fed. Cl. at 275–82. Accordingly, the court properly
determined that the record evidence rebutted the pre-
sumption of regularity.
CONCLUSION
We do not reach the Court of Federal Claims’ finding
of prejudice because the government does not contest it.
Therefore, we need not reach its argument that the Court
of Federal Claims erred in admitting the expert testimony
of Mr. Bryant Choung, which the Court of Federal Claims
relied on solely for its prejudice analysis.
We have considered the government’s remaining ar-
guments and find them unpersuasive. We affirm the
judgment of the Court of Federal Claims that the Army
must satisfy the requirements of 10 U.S.C. § 2377, which,
thus far, the Army has failed to do. Only after the Army
has complied with 10 U.S.C. § 2377 should it proceed to
award a contract to meet its DCGS-A2 requirements. To
be clear, we are not suggesting that the Army must
choose Palantir as the awardee. We simply affirm that
the Army must satisfy the requirements of 10 U.S.C.
§ 2377.
AFFIRMED
COSTS
Costs to Appellee.