Before coronavirus pandemic, parts of the intelligence community were already preparing to deal with another major disruption to the way it does business: The rise of emerging technology.
In response to the immediate threat from the coronavirus, Ellen McCarthy, the assistant secretary for the State Department’s Bureau of Intelligence and Research, said Wednesday that her organization continues with the work of gathering sensitive geopolitical information for the agency’s diplomats.
“We were prepared, we had a plan for moving analysts out of our spaces. Those analysts who are still working in our spaces are safe, and as a result, we’ve been doing a tremendous job in terms of supporting the secretary and supporting the rest of State,” McCarthy said during a webinar hosted by the Intelligence National Security Alliance.
But with the rise of artificial intelligence, geospatial data, and advanced analytics, the Bureau of Intelligence and Research, one of the smaller agencies in the national intelligence community, is now finding its niche in a more data-centric world.
“Intelligence used to be, I believe, one of the main data streams that the foreign policymaker, the diplomat, used to make policy or implement policy. I’m not so sure that happens as much today — there’s so much data that your average diplomat uses to make decisions upon which intelligence is just one stream,” McCarthy said.
While a “combination of tools and culture” prevented the intelligence community from coming up with the strategy sooner, McCarthy said INR has embraced open-source, publicly available information as the starting point for State Department intelligence.
“Some of the best data we have actually resides in publicly available information. This means that there’s going to be absolutely a hunger to actually go outside the classified domain department and try and gain those insights,” she said. “That’s not to say that — we don’t still value compartmented sensitive information, but to the extent that we can focus our resources on areas where we have gaps in knowledge, and really take advantage of what we understand from the open source, I really think that’s going to be the bright side of all of this. And I definitely get the sense that that’s what we’re going to go.”
Working with open-source information also makes it easier for INR to share its analysis, since most of its clients work at the secret level and lower, but the bureau largely operates at a higher classification level.
“How can we produce value-added intelligence that will actually help the policy they make their decisions or implement their decisions,” McCarthy said. “Because in the absence of that, they’re going to use other sources of data, and it’s not always accurate, so that’s one of our challenges.”
While a smaller IC member, McCarthy said INR has a “highly adaptive workforce” that punches above its weight. The average employee has 14 years of experience, speaks multiple languages and works alongside the diplomatic arm of the State Department to provide them with the intelligence they need.
“With those years of experience and with those relationships, they truly understand what the needs of the policymakers are, and they’re really good at what they do. They know when to deliver the intelligence, what kind of intelligence is needed, but they’re also experts in the IC and so they’re very good at leveraging other agencies to get the collections that we need or get the support that we need,” McCarthy said.
Ben Brake, the director of cyber affairs, said doing more with open-source data is just one of the ways that INR under McCarthy’s leadership has been “doing more outside of the SCIF.”
“The key to our success so far has been flexibility and everyone turning to new and other resources that perhaps our time in the SCIF didn’t allow us time to consider,” Brake said.
Still need humans ‘in the loop’
Brake said INR is keeping an eye on emerging technology not just to streamline its own processes, but also to keep track of how adversaries might take advantage of those same tools.
But for AI and data analysis tools to improve the intelligence gathering of the bureau, he said these tools must require having a “human in the loop” at all times.
“You can have really visually stunning graphics drawing information out of that, but that might not be knowledge, because different information or data sets are collected differently,” Brake said. “I think bringing the INR analyst to bear on a lot of these data sets and bring that human in the loop on the big data analysis can be really important.”
An analyst can also provide important context to the information they’re gathering – not just “what” to look out for, but “when” and “why.”
The intelligence community, back in 2008, for example, accurately predicted the “emergence of a novel, highly transmissible, and virulent human respiratory illness for which there are no adequate countermeasures” that “could initiate a global pandemic” by 2025.
But what the intelligence community didn’t have at the time, however, was a good sense of when a pandemic would hit.
“Our IC scientists were on it. They can tell us the ‘what,’ [but] the ‘when’ is the challenging part. With a lot of emerging technologies, that that is going to be a key to our analysis, when to care about it. Is this a one-to-three year commercially viable option? Is this a three-to-five 2023, or a five to never? Having that cadre of analysts on board is going to be really key to identifying when and which technologies to care about,” Brake said.