The spy service without a brake

The CIA’s deputy director now signs his own permission slips, and history shows where that leads.

(cia.gov)

The New York Times reported Monday that the career lawyer serving as the Central Intelligence Agency’s acting general counsel—installed earlier in the year—was demoted to deputy status. Michael Ellis, the deputy director of the Central Intelligence Agency, assumed the additional role of principal counsel.

The agency’s public statement frames this as a temporary measure pending Senate confirmation of Joshua Simmons, Trump’s nominee for the general counsel slot; his confirmation was scheduled for Wednesday. But the timing of Ellis’s move is unclear, and so is what legal judgments he might have made while wearing both hats.

Ellis did not need to do this. By pushing aside the acting general counsel and giving himself the authority, Ellis ensured that no independent lawyer stood between his own desk and the agency’s most sensitive decisions.

Perhaps the action was in response to the acting general counsel resisting a line of action. In recent weeks, we’ve seen U.S. attorneys threatened or fired for refusing political pressure. Without evidence, we cannot know, and speculation is easy. But even putting aside conspiracy theories, the fact remains: Ellis chose to collapse the firewall, not rely on the counsel already there.

That choice is telling. It reflects a readiness to take whatever action is necessary to get to the outcome he and his patrons want. In intelligence, the general counsel is not a legal concierge. The role exists as the mandated brake, empowered to halt missions that cross the line of law, ethics, or oversight. By making himself that lawyer, Ellis removed the possibility of hearing “no.”

I offer this not as conjecture but from experience. As executive assistant to a CIA deputy director, I saw how essential it was that the general counsel stand apart. The deputy’s role is to direct policy and operations; the counsel’s function is to apply the law to those ambitions. Their duties intersect, but their purposes diverge. One pushes forward; the other draws the legal line. Merge them, and the adherence to the rule of law collapses.

In many professions, moral brakes are weakened or ignored at cost—reputational damage, lawsuits, market loss. But in espionage—where secrecy magnifies error and lives are on the line—losing that brake can be lethal. A permissive structure ensures expedience always wins, and what began as a minor transgression compounds into systemic failure.

History shows the cost. After the Sept. 11, 2001, attacks, when counterterrorism became the CIA’s top mission, John Rizzo held the role of acting general counsel through much of the decade. He secured Justice Department signoffs before major operations—but those sign-offs came via Office of Legal Counsel opinions that defined torture so narrowly that techniques such as waterboarding, prolonged sleep deprivation, and stress positions were permitted under the label “enhanced interrogation.” He also approved the system of extraordinary rendition: capturing detainees abroad and flying them for interrogation to CIA “black sites” or third countries beyond judicial reach. He signed off on drone strikes targeting terrorists even when civilians died. None of these was a rogue act. Each carried legal cover because Rizzo endorsed them.

Rizzo’s defenders said he gave CIA officers confidence to act; his critics said he had subverted law to shield abuse. Both are true—and both show the same pathology: when the general counsel’s loyalty is recast from constitutional fidelity to policy enforcement, the office ceases to check power and becomes its stamp. That was with a separate general counsel still in place. Ellis’s maneuver goes further. If a Rizzo could bend under pressure, a deputy director who makes himself the lawyer erases the safeguard entirely.

Some might point out that Simmons, if confirmed, likely would be no check at all. But Ellis’s action is not about filling a gap until Simmons arrives. It is about showing that even the pretense of independence is unnecessary to this administration. A Senate-confirmed general counsel might well be compliant, but the office at least sustains the expectation of a separate voice. Ellis’s move discards that fiction outright.

Failures in intelligence are inevitable. Agents err, information misleads, decisions misfire. Without an independent lawyer in the room, the next failures will be larger, bolder, and harder to contain.

What should alarm us most is not the prospect of failure itself but the indifference. This administration treats independence as an obstacle, not a necessity. Oversight is brushed off as theater, accountability as inconvenience. Ellis’s maneuver reflects that ethos: The moral compass is unnecessary if you plan to follow only the map you draw yourself.

This moment should scare us. Intelligence law is being reimagined not as restraint but as permission. Remove the internal affairs unit from a police department, the judge advocate corps from the military, or the ethics board from a hospital, and everyone would call it reckless. Yet here, at the center of U.S. intelligence, that is exactly what has been done.

The degradation might not begin with grotesque abuses. It might start quietly: thinner legal opinions, fewer pauses for doubt, more risk taken without challenge. But with enough failures stacked atop one another, the outcome will be crisis—of lives lost, alliances shattered, or operations so damaging the blowback lasts decades.

Intelligence is unforgiving. In its world, law is not window dressing. It is the thin line between covert success and national disgrace.

Brian O’Neill, a retired senior executive from the CIA and National Counterterrorism Center, is an instructor on strategic intelligence at Georgia Tech. His Safehouse Briefing Substack looks at what’s ahead in global security, geopolitics, and national strategy.

---