Duty to Intervene

What It Is

The duty-to-intervene doctrine holds that an officer who witnesses another officer’s constitutional violation may be liable under 42 U.S.C. § 1983 if the officer had knowledge of the violation, a realistic opportunity to prevent or stop it, and failed to take reasonable steps. Liability attaches to the non-acting officer’s own failure – not to passive proximity, shared unit membership, or rank.

The doctrine applies to bystander officers generally, not only supervisors. Byrd v. Brishke (7th Cir. 1972) established the principle clearly: the duty flows from the officer’s role and opportunity, not from formal authority. A patrol officer watching a fellow patrol officer beat a restrained subject bears the same constitutional obligation as that subject’s sergeant.

The Standard: Knowledge Plus Realistic Opportunity

No single Supreme Court decision establishes a universal test. The doctrine developed in the federal circuits. Yang v. Hardin (7th Cir. 1994) provides the most widely cited formulation: an officer may be liable where the officer had reason to know that another officer was committing a constitutional violation and had a realistic opportunity to prevent the harm.

Knowledge is assessed by what the officer saw, heard, and could reasonably be expected to recognize – not by what the officer claims to have noticed. Prolonged force against a helpless or restrained person, continued strikes after resistance clearly ends, and repeated conducted-energy-weapon cycles after compliance are contexts where knowledge is readily inferred. A single, instantaneous shot that is over before it can be processed is a different situation.

Realistic opportunity is the doctrine’s principal limiting principle. Courts regularly deny liability where an event was too compressed to permit response. Cunningham v. Gates (9th Cir. 2000) is the leading case: officers who lacked a practical ability to intervene in rapidly unfolding gunfire were not liable merely because they were present. Presence is not opportunity.

Duration and the Sequence Problem

The most critical analytical distinction is between an instantaneous act and a continuing sequence. An officer may have no opportunity to prevent the first shot or strike. If force continues beyond the point where the threat has ended – or is repeatedly applied against a restrained or incapacitated subject – each additional act creates a new decision point, and with it, a potential new opportunity to intervene.

Lolli v. County of Orange (9th Cir. 2003) illustrates the contrast with Cunningham: officers who observed a prolonged beating up close could face liability, while officers not present had no sufficiently direct involvement. Together, the cases show both sides of the rule – rapidly unfolding gunfire may leave no realistic opportunity; an extended, visible beating may create one.

The incident should be analyzed over time, not as a single undivided event.

What Counts as Intervention

The duty does not require physically restraining another officer. Reasonable intervention includes:

• verbal commands to stop
• repositioning or assuming control of the subject
• physically separating the officers
• summoning a supervisor or additional officers
• covering or removing a weapon
• any other action reasonably calculated to prevent or reduce the harm

The obligation is to take reasonable steps – not to guarantee success. An officer who makes a genuine and reasonable attempt that fails is not liable because harm continued. Nute v. White (11th Cir. 2025) reaffirmed that this duty does not depend on supervisory authority.

Causation

The failure to intervene must cause or contribute to the constitutional injury. If the harm was complete before the bystander officer could realistically act, the failure to intervene did not cause it. Where force continues – prolonged restraint, repeated strikes, ongoing CEW cycles – intervention could have shortened its duration or severity, and causation may be established.

Edged-Weapon Encounters: Role-Specific Analysis

Multi-officer edged-weapon incidents typically involve differentiated roles: contact officer, lethal-cover officer, shield operator, CEW officer, K9 handler, supervisor, and containment officers. The duty-to-intervene analysis is role-specific, not team-wide.

First shot versus subsequent shots. Where an officer fires in response to an immediate perceived knife threat, another officer may have no realistic opportunity to prevent that shot. If the tactical circumstances change – the subject falls, drops the weapon, or loses mobility – and force continues, the analysis changes. Waterman v. Batton (4th Cir. 2005) held that each temporal phase of force must be assessed separately; Zion v. County of Orange (9th Cir. 2017) applied the same principle to an edged-weapon encounter. The intervention duty is subject to the same temporal division.

Lethal-cover officer. An officer assigned lethal overwatch who abandons that position to physically restrain a fellow officer may expose others to danger. Role assignment does not eliminate the duty – verbal commands, warnings, or supervisor notification may remain available even when physical intervention would compromise cover.

Shield operator. A shield officer is physically proximate to the subject and fellow officers but may not be able to safely release the shield to intervene. The question is what reasonable intervention was actually available, not whether physical intervention was possible in the abstract.

Disarmed or incapacitated subjects. Where a subject has been disarmed, has dropped the weapon, or has been physically incapacitated, and force continues, the opportunity for other officers to recognize and intervene increases substantially. Continued force against a person who is no longer a threat generates both an excessive-force claim and a failure-to-intervene claim for officers who observed it and had opportunity to act.

Qualified Immunity

Failure-to-intervene defendants may invoke qualified immunity. The inquiry is individualized: did this officer’s failure violate the Constitution, and was it clearly established that an officer in this officer’s specific position had a duty and realistic opportunity to act? A general proposition that excessive force must be stopped may not be specific enough in some circuits to overcome immunity in a rapidly evolving encounter. Timing, officer position, duration, visibility, and competing threats all bear on the analysis.

Policy Implications

1. Define intervention as an affirmative duty, not a permission. Policy that says officers “may” intervene is not the same as policy that says they “shall.” The constitutional doctrine is affirmative.

2. Train specific intervention behaviors. Officers need rehearsed options: verbal commands to stop, assumption of subject control, physical separation, supervisor notification, covering a weapon. The abstract duty without trained behaviors leaves officers unprepared for the moment it arises.

3. Rank should not be a practical barrier. Policy should make clear that a junior officer who witnesses unconstitutional force by a senior officer has an obligation – and is protected for acting on it. The constitutional doctrine does not vary with the chain of command.

4. Role assignment should preserve intervention capacity. Tactical structures that assign every officer to a task with no capacity to observe what other officers are doing may make intervention practically impossible. Agencies should consider whether operational designs include an officer who can recognize when force should stop and has a safe means to act.

5. Repeated force creates new decision points for every officer at the scene. Each additional strike, CEW cycle, dog bite, or continued restraint is not a continuation of the original force decision – it is a new one. Agency culture that treats continued force as a single undivided event obscures both the intervention duty and the force analysis.

6. Report who was present, what each officer observed, and what each did. Post-incident reports that document only who used force leave no record of whether the duty to intervene arose and was acted on. The resulting evidentiary gap operates against officers in litigation and against the agency in Monell analysis.

How to cite

Edged-Weapon Response Lab. (2026). Duty to Intervene. In EWRL Bulletins (Version 1.0). https://droptheknife.org/bulletins/duty_to_intervene