A few nightmare scenarios haunt the dreams of civil libertarians—scenes drawn from our long and ignominious history of intelligence abuses. One—call it the Nixon scenario—is that the machinery of the security state will fall into the hands of an autocratic executive, disdainful of the rule of law, who equates “national security” with the security of his own grip on political authority, who is all too willing to turn powers meant to protect us from foreign adversaries against his domestic political opponents, and who lacks any qualms about quashing inquiries into his own illegal conduct or that of his allies. Another—call it the Hoover scenario—is that the intelligence agencies anxious to protect their own powers and prerogatives will themselves slip the leash, using their command of embarrassing secrets to intimidate (and in extreme cases perhaps even select) their own nominal masters. As the American surveillance state has ballooned over the past 15 years, we’ve often invoked those scenarios to argue out that the slippery slope from a reasonable-sounding security measure a tool of anti-democratic repression is disquietingly short and well-oiled. You may trust that some new authority will only be used to monitor terrorists today, but under a more authoritarian administration, might it be used to suppress dissent—as when civil rights and anti-war activists became the targets of the FBI’s notorious COINTELPRO? You may be reassured by all the rigid rules and layers of oversight designed to keep the Intelligence Community accountable, but will those mechanisms function if the intelligence agencies decide to use their broad powers to cow their own overseers?

We are now, it seems, watching both scenarios play out simultaneously. Perhaps surprisingly, however, they’re playing out in opposition to each other—for the moment. Whatever the outcome of that conflict, it seems unlikely to bode well for American liberal democracy.

On the one hand we have Donald Trump, whose thin-skinned vindictiveness and contempt for judicial checks on his whims are on daily display, and who during his presidential campaign revealed a disturbing instinct for lashing out at political opponents with threats to disclose embarrassing personal information. (Recall his tweets promising to “spill the beans” on Heidi Cruz, wife of primary opponent Ted, or his warning that the Ricketts family, which funded ads opposing him, had “better be careful” because they “have a lot to hide”.) As a private citizen, Trump treated the legal system as a tool to harass people who wrote unflattering things about him; as a candidate, he thought nothing of offhandedly suggesting he could use the power of the Justice Department to jail his opponent. Even before taking the Oval Office, then, Trump had provided civil libertarians and intelligence community insiders with a rare point of consensus: Both feared that with control of both the intelligence agencies and the institutional checks on those agencies within the executive branch, Trump would fuse a disposition to abuse power with an institutionally unique ability to get away with it. On the flip side, Trump’s dismissive attitude toward the intelligence consensus that Russia had intervened to aid him in the election; his frankly bizarre, fawning posture toward Russia’s strongman leader; and his insistence on defying decades of political norms to shield his finances from public scrutiny signaled that inquiries into illicit conduct by himself or his allies and associates would be likely to wither on the vine once Trump loyalists had been installed at the heads of law enforcement agencies. As Nixon scenarios go, to steal a turn of phrase from my colleague Gene Healy, Trump is a civil libertarian’s grimmest thought experiment come to life.

And yet. 

For all that, it’s difficult not to be a bit uneasy about the way the way the national security establishment, or factions with in it, appear to be pushing back—at least, assuming the leaks that have dominated headlines in recent weeks are originating within the IC. We have witnessed the torpedoing of the president’s appointed national security adviser—by means of a decision to illegally leak the contents (or, more precisely, sources’ characterizations of the contents) of foreign intelligence intercepts of his phone conversations with the Russian ambassador. That was followed almost immediately by the explosive, albeit vague, news that—contra the administration’s denials—senior Trump associates and campaign aides had regular contact with Russian intelligence officials over the past year, though this time without any description of what those conversations concerned. 

The public interest in knowing these facts is clear enough, and under the circumstances, it is not hard to reconstruct why officials within the intelligence community might regard the drastic step of going directly to the press as necessary under extraordinary circumstances. We can infer that the ongoing investigation into the Trump campaigns Russian ties hasn’t turned up any smoking gun evidence of collusion yet, or that would likely have leaked already as well. Yet there’s presumably enough smoke that investigators are anxious to either render it politically impossible for the new administration to kill any ongoing inquiry, or—failing that—ensure that Congress feels constrained to pick up the baton after the agents working the case are reassigned to Juneau. Critically, however, this is not traditional “whistleblowing” about misconduct that a leaker has observed within their own agency, but rather disclosure of information gleaned from intelligence collection on Americans. 

That ought to raise disturbing echoes of J. Edgar Hoover’s notorious “Official and Confidential” and “Personal and Confidential” archives—troves of salacious dirt on public figures that made the FBI director a dangerous man to cross. As Hoover’s aura of omniscience grew over his three decade tenure, policymakers and even presidents were cowed by the prospect of finding their dirty laundry aired in the tabloids should they earn Hoover’s ire. Whether or not the leakers intend it, the perception that the IC is waging war on Trump is likely to resurrect that toxic chilling effect. The lesson many commentators are now drawing—some apprehensively, a few with gloating enthusiasm—is “getting on the wrong side of the Deep State can be hazardous to your political health,” which is an unhealthy notion for officials in a liberal democracy to have lodged in their heads. 

Moreover, the tension between these two scenarios is inherently unstable. “If you come at the king,” as one great political thinker has observed, “you’d best not miss,” and doubly so when the king is your employer. The New York Times recently reported that the Trump would be tapping an old business associate—who notably lacks any intelligence background—to conduct an overarching review of the intelligence community, perhaps as a prelude to a future leadership role. That has reportedly created a fair amount of anxiety in intelligence circles. Trump allies like Rep. Steve King (R‑Iowa) have already ominously suggested that “people there need to be rooted out,” and the narrative of a disloyal or hostile intelligence community could help give Trump cover to launch a purge within the agencies and install his own loyalists.

That might be the truly worst-case scenario. The career bureaucracy of the intelligence agencies, whatever its own biases and pathologies, constitutes in practice one of the few real bulwarks against the twin threats of politicized intelligence and abuse of surveillance powers. Congress, the secret FISA Court, and the IC’s Inspectors General conduct largely reactive oversight over the intelligence agencies, typically relying on internal reports of problems or some public scandal to spur them to action. Day-to-day, the primary guarantor we have that intelligence powers are being used lawfully—and that intelligence products reflect a sincere attempt to assess the truth rather than provide cover for an administration’s agenda—is the culture within the intelligence agencies, maintained largely by the middle-tier of career professionals who normally serve across multiple administrations. In what I’ve somewhat crudely called the Hoover Scenario, the intelligence establishment can become a kind of unaccountable “double government” free to serve its own interests and agendas. But that may be the lesser evil when compared with an intelligence bureaucracy that is too completely the tool of the political branches—more loyal to the president to whom they owe their careers than to the norms and mission of their agencies, and more concerned with keeping him satisfied than telling uncomfortable truths.