Robert S. Litt served in the Justice Department from 1978 to 1984 and 1994 to 1999 and as general counsel of the Office of the Director of National Intelligence from 2009 to 2016. John E. McLaughlin served as deputy director of the CIA from 2000 to 2004 and acting director in 2004 and teaches at the Johns Hopkins School of Advanced International Studies.

John Durham, the federal prosecutor chosen by Attorney General William P. Barr to examine the origins of the Russia investigation, is reportedly reviewing the intelligence community’s conclusions about Russian interference in our election. Although Durham has not confirmed the precise scope of his investigation, if these reports are correct, it is a worrisome development.

We have served in the intelligence community, in analytical, managerial and legal capacities, and in the Justice Department as a line prosecutor and at a policy level. Assuming reports of Durham’s intentions are accurate, we are profoundly disturbed that a prosecutor — even one as experienced and respected as Durham — would seek to review the conclusions of intelligence analysts.

One reason is that the mind-set prosecutors use in evaluating information is entirely different from the way intelligence analysts reach their judgments. A second is that to have the Justice Department second-guessing analysts’ work through a criminal investigation risks compromising the necessary independence and nonpartisanship of the intelligence community — particularly in a highly politicized context in which the attorney general and president appear to have prejudged the conclusion.

A prosecutor looks for documents and testimony to introduce in court that will withstand cross-examination and convince a jury beyond a reasonable doubt. An intelligence analyst works with very different material. Sometimes an analyst may have concrete proof, such as indisputable imagery or telemetry from a rocket launch — though even that requires expert interpretation. More often he or she is seeking to assess what an adversary is thinking, intending or doing based on incomplete information.

But an intelligence analyst does not have a prosecutor’s luxury to decline to proceed in the face of ambiguous information; even when information is incomplete and conclusions uncertain, policymakers need to make decisions. We will rarely, if ever, have definitive proof of what Vladimir Putin is planning, or of the state of North Korea’s nuclear arsenal, or whether Osama bin Laden is hiding in Abbottabad (estimates before the raid that killed him varied from 40 to 80 percent likely), but the president and his advisers still must decide what to do even in the absence of proof beyond a reasonable doubt. They need the intelligence community’s best understanding — even when that understanding is incomplete or inferential.

The nature of analysis is that there is almost always room for argument. Someone usually views the situation differently in whole or in part, and analytic product will typically note such dissenting views. Indeed, the intelligence community’s January 2017 assessment of Russian influence in the 2016 election explained that the CIA and FBI had high confidence that Putin and the Russian government wanted to help Donald Trump win, but that the National Security Agency had only moderate confidence in that judgment. It is impossible to imagine a prosecutor telling a jury that she has “moderate confidence” in a case, but intelligence analysts have to make those calls every day.

So a prosecutor looking at intelligence analysis must be acutely aware that analysts are dealing with information of a very different sort than prosecutors; that their goal is to reach judgments, not proof; and that the existence of some ambiguity, wiggle room or disagreement is not inconsistent with an analyst’s high confidence in a judgment.

This does not mean that analytical judgments should be free from scrutiny. Intelligence analysts, like everyone else, can make mistakes, become captives of their own preconceptions or miss significant facts — as in the case of the intelligence community’s incorrect conclusion in 2002 that Iraq had weapons of mass destruction.

But there are established mechanisms to oversee intelligence analysis — independent inspectors general in every agency who understand analytical processes, and the congressional oversight committees, the staff of which includes professionals with experience in intelligence.

Indeed, there has already been significant congressional oversight of the intelligence community’s conclusions with respect to Russia. One of us (Litt) was present at briefings the intelligence community gave to both intelligence committees, at which the conclusions about Russia’s motives were strongly challenged by certain members and equally strongly defended by career analysts. More recently, the Senate Intelligence Committee, after months of work, issued a bipartisan report concluding, as the intelligence agencies had, that Russia “sought to influence the 2016 U.S. presidential election by harming Hillary Clinton’s chances of success and supporting Donald Trump.”

Subjecting intelligence professionals to criminal investigation and the possibility of criminal prosecution for their analysis is profoundly dangerous. It is especially so when the scrutiny of their work can be seen as driven by politics — by three years’ worth of critical tweets from the president. It is a hoary cliche, but no less true, that the function of the intelligence community is to speak truth to power, regardless of how unpalatable the truth may be. If analysts fear they may come under criminal investigation for judgments the president does not like, our nation will be less safe.

Read more: