Human intelligence (HUMINT) has long been an informational cornerstone of the U.S. Intelligence Community (IC), providing crucial insights into the intentions and actions of foreign adversaries. Yet, as highlighted in a recent Washington Post article, the CIA's ability to recruit and leverage human sources abroad may be facing significant challenges. This reported decline in HUMINT capabilities has significant implications for U.S. national security.
In an era where traditional espionage methods are increasingly constrained, the United States must adapt. One key tool may be the integration of probabilistic forecasting into existing open-source intelligence (OSINT) methods, which could help mitigate the gaps left by diminished HUMINT capabilities. Both methods have a proven track records that could help the United States monitor, predict, plan, and respond to global strategic developments.
OSINT is no longer a peripheral tool; it is rapidly becoming a central pillar of modern intelligence. RAND research has consistently highlighted OSINT's growing relevance, driven by the increasing availability of open-source data—ranging from social media feeds to commercial datasets. These data sources, once considered secondary or tertiary sources of key information, may now offer critical insights into adversaries' activities, public sentiment, and emerging trends.
Technological advancements are further enhancing OSINT's utility. Machine learning and natural language processing tools are revolutionizing the ability to sort, translate, and analyze vast amounts of data. For example, algorithms can now identify patterns in social media activity, detect disinformation campaigns, and track the movement of goods and personnel in near real-time. These capabilities make OSINT a valuable resource for understanding the global security landscape.
However, OSINT remains underutilized by the IC, largely due to gaps in tradecraft and institutional policies. Unlike traditional intelligence disciplines such as HUMINT, Geospatial Intelligence, or Signals Intelligence, OSINT lacks a standardized framework for assessing source credibility and integrating data into analytic products. This reportedly has led to skepticism about its reliability and limited its adoption within the IC.
One solution to these challenges lies in the concept of “third-generation OSINT.” As RAND research has found, this next iteration of OSINT will be built on automated reasoning and machine learning, enabling more efficient collection and dissemination of intelligence. It may also emphasize crowd-sourced forecasting, a methodology that leverages collective expertise to predict future events.
Probabilistic forecasting is a data-driven technique designed to assess the likelihood of future events, such as geopolitical conflicts, economic shifts, or discrete military actions. Forecasting models can generate actionable insights that help policymakers anticipate and prepare for emerging threats by analyzing patterns in historical data.
Forecasting models can generate actionable insights that help policymakers anticipate and prepare for emerging threats by analyzing patterns in historical data.
There are vast amounts of public data available through open sources that can provide the foundation for probabilistic models, enabling forecasters to identify risks and opportunities with greater accuracy. Moreover, forecasting addresses one of the most common criticisms of OSINT—its perceived lack of credibility. Probabilistic forecasting could help to ensure that intelligence products are grounded in evidence and transparent reasoning (PDF).
RAND's Forecasting Initiative (RFI) has successfully demonstrated the power of forecasting in addressing complex challenges. For instance, probabilistic models have been used to monitor Russia's disinformation campaigns, assess the risk of pandemics being weaponized and tracking political instability in Africa. These initial insights, which are informing ongoing RAND research, highlight the value of forecasting as a tool for strategic decisionmaking. While forecasting is not a substitute for traditional intelligence collection methods, it can bolster U.S. intelligence efforts as the nation seeks to rebuild its HUMINT capabilities and capacity.
The decline in HUMINT capabilities likely necessitates a shift in intelligence collection sources. Probabilistic crowdsourced forecasting could offer a way to mitigate collection gaps and complement traditional methods by combining human expertise and judgments with data-driven insights.
Institutional change and cultural acceptance will be essential to fully realize the potential of these tools. The United States must develop robust tradecraft methodologies for OSINT analysts, ensuring that it has the skills and frameworks needed to assess credibility, analyze data, and produce high-quality intelligence products. As traditional espionage becomes more challenging, the United States must embrace these modern methodologies to maintain its edge in a rapidly changing world.
