Palantir CEO Alex Karp has warned that advances in Super Intelligence (SI) could eventually allow systems to predict human thoughts and actions, raising concerns about how governments and private companies might use the technology to monitor or influence people.

In an interview with Louis Robson, Karp discussed the possibility of predictive systems extending beyond military operations into domestic life. The technology builds on algorithms already used by the U.S. Department of Defense and its allies to analyze large volumes of information and anticipate potential developments.

On the battlefield, these systems can process troop movements, communications activity and satellite imagery to help forecast an adversary’s next moves. Similar predictive techniques could be applied domestically by analyzing consumer habits, digital communications, location data and biometric information to estimate how individuals might behave.

The prospect raises questions about privacy and individual freedom, particularly if predictive systems are deployed without meaningful consent or oversight. Such capabilities could give governments or corporations the ability to anticipate decisions and potentially influence people’s choices before they act.

Karp also addressed competition among major technology companies, including OpenAI and Anthropic, warning about the possibility of control over advanced systems becoming concentrated in a small number of powerful firms. He has argued for technological sovereignty and decentralized governance, with local institutions maintaining greater control over how the technology is developed and used.

He has also criticized consumption-based business models that rely on extracting value from companies’ private data and intellectual property. Karp has suggested that the risks and liabilities associated with advanced technology could eventually prompt calls for greater government intervention, including nationalization of some technology operations.

The debate also highlights tensions within the technology industry over the potential dangers of increasingly capable systems. Karp has historically maintained that Palantir’s defense software was intended to support operations against foreign adversaries rather than enable mass domestic surveillance of Americans. His warnings about predictive capabilities in domestic settings raise questions about how those boundaries could be maintained as the technology advances.

Meanwhile, Palantir Chief Technology Officer Shyam Sankar has dismissed extreme warnings about catastrophic SI scenarios as “mass psychosis,” underscoring disagreements within the industry over the severity of potential risks.

Although the ability to reliably predict an individual’s thoughts or future actions remains an uncertain prospect, the growing use of data-driven behavioral analysis has already raised concerns about surveillance, privacy and the concentration of technological power. The central question is whether effective safeguards can keep pace with increasingly sophisticated predictive systems.