Two Ways to Stop Secrecy from Undermining US National Security

The U.S. system of classification protects the sources of intelligence and the methods by which it is collected, but it also prevents American allies and partners from seeing all but a trickle of the data and analysis produced by the U.S. intelligence community. This is no longer a tenable approach in an era where information is key to exposing and countering aggressive, dangerous and deceptive actions by China, Russia, and others. 

Whether it is the fight to combat Russian disinformation and election interference, the challenge of highlighting predatory loans associated with Beijing’s Belt and Road Initiative, or simply convincing potential partners to cooperate with Western nations, U.S. officials must routinely soft-pedal threats or remain excessively vague during international discussions. This situation makes it very difficult to achieve the goals outlined in the National Security Strategy and National Defense Strategy, especially deterring adversaries while strengthening alliances and building partnerships.

A recent memo to the Director of National Intelligence signed by senior military officers and endorsed by the Defense Undersecretary for Intelligence and Security calls for allowing more data and intelligence to be shared with allies and partners. The sentiment has also been echoed by other current and former senior officials within DOD and the Intelligence Community.

The solution has two parts: change how the foreign policy and national security communities collect and analyze data; and shift how they view acceptable risk.

The easiest way to begin producing more releasable information is to make more and better use of information that was never classified in the first place. This publicly available information, or PAI, includes news, academic research, financial transactions, social media, blogs, advertising data, dark web searches—and the new sources and types of information that appear every day in this era of Big Data. 

To be sure, PAI must be sifted and evaluated, just like any classified information, to find the credible and useful bits. The intelligence products and planning documents built on PAI must meet the same rigorous analytical standards as those derived from traditional intelligence methods. 

PAI supplies the ingredients of open source intelligence (OSINT), which in turn, should become the foundation for the vast majority of intelligence products. This will allow signals intelligence, satellite imagery, and human reporting to be reserved for the hardest, most inaccessible targets. It also sets up a way to prove or validate assessments made using classified sources (and vice versa). This validation deepens policymakers’ trust in the analysis and can help leaders ensure resources are being used properly.

More importantly, using OSINT to prove classified data allows the U.S. government to increase the releasability of information. For instance, the intelligence community might be aware of nefarious activity in a partner nation through signals intelligence, but this information cannot be readily shared because of the method of collection. If U.S. analysts use PAI to arrive at that same conclusion, however, that information can be shared with those who need to know, and it will be supported purely by unclassified information.

Making more use of PAI would also help the IC become more flexible. Amid the COVID-19 pandemic, several intelligence agencies are exploring remote work and concluding that OSINT is a necessary part of the solution.

Another way to increase shareable information is to increase participation in coalitions and combined operations with partners, whose information products tend to be more releasable. NATO represents the best example of such efforts, but other valuable initiatives include the maritime-security-focused Information Fusion Centre hosted by Singapore. For example, coalition or information sharing agreement in the Indo-Pacific would allow information about China’s malign activities to be better shared, which would help combat Beijing’s narrative of being a benign power. In general, the Defense Department and IC should look for more opportunities to work with other militaries and spy services.

Certainly, there are times when traditional intelligence and analysis methods will be the only way for foreign policy and security leaders to know certain information. The solution here is for national security policymakers to accept more risk in releasing information to international partners who need to know. There is a risk that intelligence operations or methods might be exposed or compromised, but there is an alternative risk that failing to share enough information will prevent the United States from achieving its strategic objectives and leave allies exposed to danger. 

These risks must be recognized and weighed—and that requires a fundamental relook at the goals and processes of classification. Current processes are overly restrictive and not designed for a modern, data-driven world. Commonly known facts are often classified, sometimes even at the Top Secret or compartmented levels, and requests to share information are often denied because of vague fears that the information could indicate the use of a satellite or a human source. This paranoid and draconian mentality is counterproductive and contrasts greatly with some of our closest allies who seek to produce mostly releasable information and only restrict the most sensitive intelligence.

Some parts of the U.S. security apparatus have begun shifting their thinking. In light of Chinese influence on college campuses and attempts to steal intellectual property, the FBI started hosting town halls and seminars highlighting counterintelligence threats. Such briefings would have been anathema to the Bureau even a few years ago, but officials realized public awareness of Chinese actions far outweighs traditional notions of secrecy and security.

Increasing the releasability of U.S. security assessments and products is not a matter of transparency. Rather, it is a matter of usefulness. Information should still be appropriately protected and only shared with necessary parties, but the government should not think only Americans need to know about today’s challenges and tomorrow’s threats.