Ahead of the the 2024 US elections, the US intelligence community and law enforcement were on high alert and ready to share information—both among agencies and publicly—as foreign malign influence operations emerged. Tech giants like Microsoft similarly sprang into action, collaborating with government partners and publishing their own information about election-related disinformation campaigns. The speed and certainty with which authorities were able to pin these efforts on threat actors in Russia, China, and Iran was unprecedented. But researchers also caution that not all attributions are created equal.

At the Cyberwarcon security conference in Arlington, Virginia, today, researchers from the Atlantic Council’s Digital Forensic Research Lab are presenting initial findings on the role of attribution in the 2024 US elections. Their research compares the impact of quickly naming and shaming foreign influence actors to other recent US elections in which government attribution was far less common.

“We’re building on a project that we did back in 2020 where there was a lot more context of concern that the Trump administration was not being forthcoming about foreign attacks,” says Emerson Brooking, director of strategy and resident senior fellow for DFRLab. “In contrast to 2020, now there was an abundance of claims by the US government of influence operations being conducted by different adversaries. So in thinking through the policy of attribution, we wanted to look at the question of overcorrection.”

In the lead-up to the 2016 US presidential election, Russia’s extensive influence operations—which included hack-and-leak campaigns as well as strategic disinformation—caught the US government by surprise. Law enforcement and the intelligence community were largely aware of Russia’s digital probing, but they didn’t have an extreme sense of urgency, and the big picture of how such activity could impact public discourse hadn’t yet come into view. After Russia’s hack of the Democratic National Committee in June that year, it took four months for the US Office of the Director of National Intelligence and the Department of Homeland Security to publicly attribute the attack to the Kremlin. Some officials had said in the weeks following the incident that formal confirmation from the US government might never come.

Even in the highly politicized landscape that followed, federal, state, and local collaboration around election security expanded dramatically. By 2020, the researchers say, 33 of the 84 influence operation attributions they studied related to the 2020 US elections, or about 39 percent, came from US intelligence or federal sources. And this year, 40 of the 80 the group tracked came from the US government. DFRLabs resident fellow Dina Sadek notes, though, that one important factor in assessing the utility of US government attributions is the quality of the information provided. The substance and specificity of the information, she says, is important to how the public views the objectivity and credibility of the statement.

Specific information confirming that Russia had manufactured a video that purported to show ballots being destroyed in Bucks County, Pennsylvania was a high-quality, useful attribution, the researchers say, because it was direct, narrow in scope, and came very quickly to minimize speculation and doubt. Repeated statements from the Office of the Director of National Intelligence’s Foreign Malign Influence Center warning very broadly and generally about Russian influence operations is an example of the type of attribution that can be less helpful, and even serve to amplify campaigns that otherwise might not register with the public at all.

Similarly, in the lead-up to the 2020 elections, the researchers point out, statements from the US government about Russia, China, and Iran playing a role in Black Lives Matter protests may have been mismatched to the moment because they didn’t include details on the extent of the activity or the specific objectives of the actors.

Even with all of this in mind, though, the researchers note that there was valuable progress in the 2024 election cycle. But with a new Trump administration coming into the White House, such transparency could start to trend in a different direction.

“We don’t want to come across like rearranging deck chairs on the Titanic, because the state of affairs that was is not the state of affairs that will be,” Brooking says. “And from a public interest perspective I think we got a lot closer on disclosure in 2024.”