A refresher on the most common misconceptions about Palantir, what we do, and how we work
Editor’s Note: This is the seventh post in Palantir Explained, a series that explores a range of topics, including our approach to privacy, security, AI/ML safety, and more. In this post, we revisit some of the themes discussed in our first installment from 2020, addressing recurrent misconceptions about who we are and we what we do. While the intervening years have witnessed many considerable disruptions, changes, and events, one point of consistency is that Palantir is still not a data company.
Who We Are and What We Do
Palantir makes software which organizations use to better manage their data, improve their operations, and serve the people who rely on them.
We’re proud that essential organizations — including those delivering life-saving assistance, improving health outcomes, manufacturing aircraft fleets, and securing and defending the West — depend on our software platforms to deliver their most vital mission outcomes and further institutional trust within the communities they serve.
Contrary to some media reports, we are not a surveillance company. We do not sell personal data of any kind. We don’t provide data-mining as a service. Palantir is a software company. Unlike many technology companies, our business model is not based on monetizing personal data. Instead, we develop and license software platforms that enable our customers to integrate and analyze their own data assets to make better decisions. Privacy and data security are fundamental to Palantir and have been built into the software’s architecture from the start.
We make digital infrastructure that enables organizations to operate in complex data environments. We help our customers — across the public, private, and non-profit sectors — overcome common challenges associated with fractured data landscapes, in which their data is split across different systems and formats.
Our software provides our customers with the capabilities to integrate those sources into a common platform in which they can build more effective data management, analytics, and operations. Many of our customers also use our platforms to build or deploy AI tools to further enhance their operations in responsible, reliable, and impactful ways.
Palantir has a deep and longstanding commitment to protecting privacy and civil liberties. We were the first company to establish a dedicated Privacy & Civil Liberties Engineering Team over a decade ago, and we have a longstanding Council of Advisors on Privacy & Civil Liberties comprised of leading experts and advocates. These functions sit at the heart of the company and help us to embody Palantir’s values both through providing rights-protective technologies and fostering a culture of responsibility around their development and use.
Transparency and Engagement
We strive to be transparent about the security and privacy controls in the technologies that we design, build, and deploy, publishing blogs, videos, documentation, and engaging at events. Anyone can sign up for our AI platform, Palantir AIP, and test it out for themselves. In fact — we encourage this to see firsthand how our software works!
Palantir’s software is built at every stage to uphold, not undermine, legal and regulatory protections as well as the ethics and standards that help institutions govern the appropriate uses of powerful technologies. A key part of this is oversight. Our software features, among other complementary capabilities, audit logs and transparent data flows, which enable regulators and governance bodies to ensure that the software has been used legally and responsibly.
In fact, it is the strength of these protections that has underpinned our success in the age of AI. These integral safeguards help to ensure that AI models can be used in safe, controlled, tested, and privacy-protective ways. Wherever possible, we strive to provide tools which exceed the protections that the law requires.
Our Longstanding Support of the U.S. Government
We are extremely proud to support key American institutions. These include longstanding contracts with agencies like the IRS and NIH, where we have a history of excellence and the efficient delivery of essential digital infrastructure. Central to our mission is supporting liberal democratic societies and their allies by providing vital institutions with software that enables them to fulfill their responsibilities and maintain the trust of the citizens they serve — a commitment that transcends political shifts and administrations.
Each of these customers gets a unique instance of our software, over which they have control. They define which data is integrated, who has access to it, and how it is used. Palantir does not have unfettered access to data our customer provide us. Nor does it share, transfer, monetize, or otherwise use such data for its own business purposes.
If any Palantir engineer has temporary access to data, this is at the explicit instruction of the customer, who grants individual access to their instance of the software, and assigns the permissions necessary to carry out discrete technical tasks. This access is then revoked once the project is complete.
Enabling Secure Collaboration Between Organizations
Many institutions — including U.S. federal government agencies — rely on our software because it helps them deliver effective and reliable outcomes. However, the fact that two agencies run on the same software does not imply generalized data sharing between them (just as the fact that two agencies both use spreadsheets does not allow them to access each others spreadsheets).
Where limited sharing or collaboration between agencies does need to take place, this is always directed by the agencies in question, for a specific purpose, explicitly in accordance in accordance with our legal obligations, and protected by deep policy and technical safeguards. This is done to better meet the service obligations of the people for whom these institutions are responsible.
Misconceptions and Incorrect Reporting
Misconceptions can arise because our products are complicated, used in a huge variety of industries (often with confidentiality requirements), and support nuanced and technical functions. This category of software is unfamiliar to the average person. However, it is absolutely possible to accurately describe it to people who are curious.
The relative complexity of software is not an excuse for critics to trade on people’s fears with false or misleading claims. These organizations have a responsibility to inform people and to allow them to make their own intelligent opinions. To support this, we are always willing to engage with the media and provide clarifications on any points of uncertainty — we just ask that these organizations seek the truth in good faith.
Data integration and the modernization of software is an essential process that every functional institution must go through. The effectiveness of our democratic institutions — and therefore the upholding of trust in democracy itself — depends on it. We’ve been helping key organizations do this for more than two decades.
If this mission excites you, join us! Apply here today: palantir.com/careers
Palantir Is Still Not a Data Company (Palantir Explained, #7) was originally published in Palantir Blog on Medium, where people are continuing the conversation by highlighting and responding to this story.