One of the biggest beneficiaries of the expansion of the homeland-security industrial complex since the second inauguration of Donald Trump has been Palantir.  Shares of Palantir stock have doubled in value since Trump’s re-election.

Both the Department of Defense and the Department of Homeland Security have expanded their contracts with Palantir for data aggregation, data mining, algorithmic profiling, predictive “pre-crime” policing and preemptive war, and automated decision-making.

But is Palantir just doing more of what it has been doing since at least the first Trump presidency? Or is it (also) doing something new? We think it’s doing both.

Palantir is one of the prime contractors being paid to carry out President Trump’s executive order for the integration, mining, and use for decision-making throughout the Federal government of information about individuals held by any Federal agency, regardless of what agency originally collected it or for what purpose. Trump’s executive order seeks to define “purpose limitation”  — one of the fundamental principles of fair information practices — out of existence, at least as applied to the Federal government.

Working with and for the Department of Government Efficiency, Palantir has been central to this Federal government-wide effort to abolish “data silos”. Palantir is reportedly building aggregated databases and platforms for analysis and decision-making about both immigrants and foreign visitors and US citizens.

The expansion of Palantir’s activities has, unsurprisingly, made Palantir a focus of renewed protest.

Some of Palantir’s expansion is just more of more of what it was already doing. In particular, Palantir pioneered natural-language queries for mining of complex datasets and complex algorithms for identification of patterns in data long before either of those processes came to be labeled “artificial intelligence”. Now it’s applying these tools to a wider range of data and decisions. But the fundamental dangers remain the same. As the algorithms and the volume of data ingested become sufficiently large and complex, it becomes impossible to attribute a decision to any specific item of data or rule, or to exercise human oversight or judicial review of that algorithmic decision.

Meanwhile, Palantir’s expanded work for the US government has broken new ground, or broken through barriers, in several ways:

• Expansion from the Departments of Defense (DOD) and Homeland Security (DHS) to all components of the Federal government;
• Expansion from people and activities with some foreigners or foreign travel or trade to all US citizens; and
• Expansion from decisions about air travel and entry to the US to all types of Federal government decisions about what individuals are or aren’t allowed to do.

The impact of all of these changes is to normalize pervasive suspicionless surveillance — collection, retention, and integration of data about movements, activities, and transactions — and permission-based extrajudicial government control as the defaults.

While this is an expansion of previous government intrusions on individual freedom of movement and action,  it’s also a fundamental conceptual shift from the assumption of an “airport exception” to the US Constitution or a “non-US citizen exception” to Constitutional or human rights, to a permission-based regime of government surveillance and control applied to all individuals and all activities within reach of US government power.

In the conceptual framework that underlies Trump’s executive order on “data silos” and Palantir’s work to build an omniscient and omnipotent “artificial intelligence” platform, there are no limits to the scope of individuals or activities to which it is applied.

The DHS data lake is now a US government-wide data ocean in which we all swim, all the time, and in which Palantir is constantly monitoring and choosing which fish to corral or catch.