Anti Fraud

Palantir Anti-Fraud enables commercial enterprises and government organizations to detect and eliminate sophisticated criminal activity, including credit card bust out fraud, money laundering, check kiting, mortgage fraud, tax fraud, tax evasion, and synthetic identity fraud.

Who Needs Our Help?

Losses due to fraudulent activity are staggeringly large—more than $600 billion per year in the U.S. alone. Crime of this magnitude erodes trust in public and private institutions and, at its worst, can threaten an organization’s solvency.

Fraud continues to grow at an astonishing rate as criminals refine their tactics in response to anti-fraud measures.

The highly adaptive nature of these criminals renders industry-standard automated solutions less effective with each passing day. Organizations need a better way to interact with the massive amounts of information they collect in order to identify and eradicate patterns of fraudulent activity hidden within their enterprise data. They need an immune system for the enterprise.

Build up immunity to fraudsters with computer-assisted, human-driven analysis

Fraudsters are sophisticated criminals, changing their tactics nearly every day. To combat this kind of threat, Palantir Anti-Fraud augments human analytic capabilities by capitalizing on the strengths of machine analysis while avoiding the risks of purely automated approaches. Palantir’s approach to this kind of human-computer symbiosis was inspired by lessons learned from combating adaptive threats at PayPal. Palantir Anti-Fraud uses algorithms to comb through existing archives and streams of new data in order to isolate and surface patterns based on rules defined by human analysts. The analysts, in turn, investigate cases using their intuition, experience, and domain knowledge, identifying and responding to subtle cues, edge cases, and new indicators of evolving fraud tactics.

Identify novel patterns of fraud by intuitively analyzing data at massive scale

Palantir Anti-Fraud allows human experts to look across their entire universe of data to find novel patterns of suspicious activity. Investigators at commercial enterprises have the contextual knowledge to know where to look for unique strains of fraud, such as money laundering, check kiting, or complex, synthetic identity fraud across lines of business. Tax authorities have the experience and domain expertise to develop hypotheses about new forms of tax fraud, illegal transfer pricing, and other forms of tax evasion. Palantir Anti-Fraud’s extensible analytical applications enable these domain experts to interact with the data in ways explicitly designed for fighting fraud. With transactions, weblogs, network traffic, and other dense, low-signal, disparate data from across the enterprise fused into a coherent object model, analysts can ask questions in the language of entities, events, and relationships, not data primitives.

Eradicate known patterns of fraud across the enterprise

Once an analyst has identified and characterized a new pattern of fraudulent behavior, Palantir Anti-Fraud can quickly recognize all cases that conform to this pattern, enabling managers to take swift action to eliminate the threat. Analysts can build new rules around complex attributes or behaviors that they have discovered in the course of their investigations. The software can then run clustering algorithms persistently against the data to identify criminal behavior at massive scale, effectively creating a resistance within the enterprise to particular strains of fraud or crime.