Security Guard Use of Force: The Law, Limits & Liability (2026)

If you are hiring security for a store, apartment community, hospital, warehouse, or event, the single question that generates the most litigation is not "will the guard show up" — it is "what happens when the guard puts hands on someone." Getting this wrong is expensive: excessive-force and false-imprisonment claims routinely name not just the guard and the security company but the property owner or business that hired them. This guide explains how the law generally treats private-security force in the United States, in plain terms, so you can vet a provider before you sign. It is general education, not legal advice — statutes and case law differ meaningfully by state, and a short comparison table below shows exactly how much they can differ.

Guards are not police: force flows from private-citizen law

The foundational concept is simple and frequently misunderstood by clients and guards alike. A private security officer is a private citizen, not a government agent. Their authority to act on your premises is delegated by you, the property owner, under common-law rights that any citizen has — self-defense, defense of others, and a limited right to protect property. It is not derived from the state's police power.

That distinction has hard consequences. A guard does not have general arrest powers, cannot conduct investigatory stops or searches the way police can, and, unlike a police officer, generally does not enjoy qualified immunity. When a police officer's use of force is challenged, courts weigh it against constitutional standards with immunity protections. When a private guard uses force, they are typically judged under ordinary state tort and criminal law — assault, battery, false imprisonment — with no immunity shield (the Supreme Court declined to extend qualified immunity to privately employed guards in Richardson v. McKnight, 521 U.S. 399 (1997)). One important exception: where a guard is an off-duty sworn officer or a commissioned "special officer" acting under color of law, the analysis can flip toward state-action and immunity questions, and federal constitutional claims under 42 U.S.C. § 1983 may come into play — a scenario covered in its own section below. A guard's delegated authority also generally ends at the property line. For a deeper look at what officers can and cannot do at the point of contact, see our guide on whether security guards have arrest powers.

The core standard: reasonable and proportional force

Across jurisdictions, the legal test for any use of force is remarkably consistent. Force must be lawful, necessary, reasonable, and proportional to the threat as it reasonably appeared at the moment. Courts and juries look at three things: was the threat imminent, was force actually necessary, and was the officer's belief objectively reasonable given what they perceived at the time.

Three principles do most of the work:

• Necessity. Force is a response to a genuine threat, not a tool for control or punishment. Verbal abuse, insults, or refusal to cooperate do not, by themselves, justify hands-on force.
• Proportionality. The level of force must match the level of threat. A shove does not justify a strike; a strike does not justify a weapon.
• Duration. Force must be defensive rather than offensive, and it must cease the instant the threat is neutralized, the person complies, or the situation is under control. Continuing to apply force after that point is where many "reasonable" incidents turn into batteries.

A worked example: walking the force ladder

Abstract principles are hard to price. Here is the same encounter escalating three different ways, and where "reasonable" tips into "battery":

• Presence and verbal (the subject cooperates or simply leaves). A guard notices a shopper concealing merchandise and approaches. Standing at a professional distance and saying, "Sir, I need you to step over here" is presence plus verbal direction — no force, no exposure. If the person stops, this is where a well-run program ends the contact and lets police or a manager handle it.
• Soft control (the subject pulls away). The person turns to leave and the guard takes a light grip on an arm to keep them from exiting. A brief, non-injurious hold to maintain a lawful detention is generally within "reasonable" — if the detention itself is lawful (see shopkeeper's privilege below). Pull-away, by itself, is not an assault on the guard and does not unlock striking force.
• Hard control (the subject throws a punch). Now the person swings at the guard's head. The subject has introduced a threat of physical harm, so the guard may use hard-control techniques — blocking, a takedown, a restraint hold — reasonably calibrated to stop the assault. This is defense-of-self territory, the guard's strongest footing.
• Defensive deadly force (the subject draws a weapon). If the person produces a knife or gun and presents an imminent threat of death or serious bodily injury, the deadly-force standard may be reached. This is the extreme rung and it invites intense after-the-fact scrutiny.
• The litigated moment — force after compliance. Suppose after the takedown the subject stops resisting, goes limp, and says "okay, okay." If the guard keeps striking, keeps kneeling on the person's back, or keeps cranking a hold, the justification has already expired — the threat is over. This single fact pattern, continued force after compliance, is the most heavily litigated in use-of-force cases and the most common way a defensible incident becomes a six-figure claim.

Professional providers train officers on exactly these gradations so a guard and their supervisor can later articulate why each level was reasonable and necessary. The critical legal question is always that articulation, not a rigid checklist.

Duty to retreat vs. stand-your-ground on post

Self-defense is a guard's strongest footing, but states split on whether a person must retreat before using force. In "stand-your-ground" states, a person who is lawfully present and not the aggressor generally has no duty to retreat before using proportional force, including deadly force where the threshold is met. In "duty-to-retreat" states, a person must use a safe avenue of escape if one is available before resorting to deadly force (the "castle doctrine" typically removes that duty inside one's own home, but a commercial post is not the guard's home). For a guard, this matters because retreating, creating distance, and summoning support is not just tactically smart — in duty-to-retreat states it can be legally required before force is justified. Confirm which rule your state follows and make sure a provider's post orders reflect it.

Self-defense, defense of others, and the limited defense-of-property doctrine

A guard's strongest legal footing is self-defense and defense of others. When the officer reasonably believes they or a third person face an imminent threat of unlawful physical harm, they may use reasonable, proportional force to stop it. The threat must be actual and immediate — a speculative or future threat does not qualify.

Defense of property is far weaker. In most states a guard may use reasonable, non-deadly force to prevent trespass, theft, or damage — asking someone to leave, blocking a doorway, physically interposing to stop merchandise from walking out. But the ceiling is much lower than in self-defense. In the large majority of states, deadly force may not be used to protect property alone. A minority of states carve out narrow exceptions: Texas, for example, permits deadly force to protect property under Tex. Penal Code § 9.42, but only to prevent an imminent arson, burglary, robbery, aggravated robbery, theft during the nighttime, or criminal mischief during the nighttime — or to stop a perpetrator fleeing with property after such a crime — and only when the actor reasonably believes the property cannot be protected or recovered by other means (or that using lesser force would expose them to death or serious injury). Louisiana and a few other states have analogous carve-outs. These are narrow, heavily scrutinized, and fact-dependent; they are not a general license to shoot over goods. The law's default remains that human life outweighs property — letting a laptop or a car walk away is legally preferable to a lethal response.

For most commercial sites the practical lesson is the same regardless of your state: the right posture is deterrence and documentation, not physical recovery. That reality shapes which service you actually need — often visible presence such as mobile patrol or video surveillance deters loss more effectively, and with far less liability, than a guard tackling a shoplifter.

Force during a citizen's / private-person arrest and detention

Guards make detentions under two overlapping doctrines: the common-law citizen's (private-person) arrest and, in retail, the statutory shopkeeper's (merchant's) privilege.

Under citizen's-arrest rules, a private person may generally detain someone for a felony where there are reasonable grounds to believe that person committed it, or for a misdemeanor (in many states, a breach of the peace) committed in their presence. But there is a trap that a buyer must understand, because it is the sharpest way private authority differs from police authority. In many states the common-law rule still in force is the "felony-in-fact" requirement: a private person is justified only if a felony was in fact committed and there were reasonable grounds to believe this particular person did it. The reasonable-grounds latitude covers only who committed the crime — not whether any crime occurred. If it turns out no felony actually happened, the arrest is unlawful and the guard, the firm, and the client can be liable for false imprisonment even if the belief was entirely reasonable. A police officer who reasonably but mistakenly believes a felony occurred is generally protected; a private citizen, in these states, acts "at his peril." California codifies exactly this in Penal Code § 837 (arrest permitted "when a felony has been in fact committed, and he has reasonable cause for believing the person arrested to have committed it"), and New York's CPL § 140.30 uses the same "in fact" language. Some states have relaxed the rule, but you should assume the strict version applies unless your state's law says otherwise. Bare suspicion is never enough — especially for a misdemeanor — and the person must be turned over to police without unnecessary delay.

The shopkeeper's privilege, codified by statute in most states, softens this for retail theft: it lets a merchant and its agents detain a suspected shoplifter when three conditions are met: (1) reasonable or probable grounds to suspect theft, (2) a reasonable manner of detention, and (3) a reasonable duration limited to investigating or awaiting police. No statute sets a hard clock — "reasonable" is judged case-by-case, and courts have found even fifteen minutes too long where a quick question to a cashier would have resolved it — but long holds routinely cross into false imprisonment. The privilege is for investigation only: it does not authorize forced confessions, and it generally does not permit searching a person's clothing without consent (California's PC 490.5, for instance, limits any search to bags and packages, not the person, and lets a merchant request but not demand ID). Retail buyers should read our retail loss-prevention security guide, because well-run programs deliberately train observe-and-report over apprehension.

How the rules actually differ by state

"Varies by state" is only useful if you can see how. The table below contrasts five representative states on the three questions a multi-site buyer cares about most. Verify the current statute for your own jurisdiction before relying on any row — these are summaries, not legal advice.

The pattern is worth internalizing: the citizen's-arrest threshold ranges from strict "felony-in-fact" (CA, NY) to a broader "reasonable grounds" for any ongoing offense (IL); deadly-force-for-property is an outlier permitted only in a handful of states (TX); and even where a shopkeeper's privilege exists everywhere, its guardrails on duration and search differ. A national provider's post orders should be written to the strictest state you operate in.

Deadly force — the high bar and why armed posts raise everyone's liability

Deadly force by a private guard is reserved for one narrow situation: an imminent threat of death or serious bodily injury to the officer or another person. It is a last resort, subject to intense after-the-fact scrutiny, and (outside the narrow property carve-outs like Texas's § 9.42 discussed above) it is not lawful merely to protect property or to effect an arrest of a non-dangerous suspect.

Putting a firearm on a post changes the entire risk profile of the assignment. Armed officers face additional state licensing, firearms qualification, and screening requirements, and any deployment invites litigation. Standard business insurance often excludes firearm activities, so armed work typically requires specific endorsements (more on that below). None of this means armed guards are wrong for every site — a cash-handling or high-threat environment may need them — but the decision should be deliberate. Weigh it with our comparison of armed vs. unarmed security guards, review the armed guard requirements by state, and if you do proceed, understand what armed security actually entails and costs before you post it.

Non-firearm weapons: OC spray, batons, and TASERs are each their own legal question

Many "unarmed" posts are not truly weapon-free — officers may carry OC (pepper) spray, an expandable baton, or a conducted-energy device such as a TASER. Each of these intermediate weapons is separately regulated, and treating them as a footnote is a mistake. Depending on the state, possession or carry may require its own permit or training certification; some jurisdictions restrict baton or stun-device carry entirely, or condition it on a guard license endorsement. Each tool also carries distinct tort and medical exposure: OC spray raises decontamination and aftercare duties and can be dangerous to people with respiratory conditions; batons deliver blunt-force injury and are easy to deploy disproportionately; TASER deployments have been associated with falls, cardiac events, and in-custody deaths, and generate their own body of litigation. From an insurance standpoint, every one of these is a separate underwriting question — a policy may cover unarmed guard operations but sublimit or exclude weapon-related claims. Before a provider posts any weapon on your site, get in writing (a) that state law and the guard's license permit it, (b) the specific training and requalification standard, and (c) that your COI covers the associated liability. Compare programs against training requirements by state.

Restraint holds, positional asphyxia, and in-custody medical risk

Off-duty police and "special officers": when the clean binary breaks

The rule that a guard is "just a private citizen" has a large and important exception. A significant share of contract security is performed by moonlighting sworn officers or by commissioned "special police" whom a state or city has deputized with limited law-enforcement powers. When such an officer acts under color of law — flashing a badge, invoking their police authority, making an arrest as an officer rather than as a private employee — a court may find their conduct constitutes state action. That flips the analysis the rest of this article describes: the plaintiff can bring a federal constitutional claim under 42 U.S.C. § 1983, and qualified-immunity questions re-enter the picture (unlike the purely private guard in Richardson v. McKnight). Courts look at whether the officer was performing a public police function or merely private security work; the line is fact-intensive and litigated hard.

For you as the buyer, this cuts both ways and mostly toward more exposure, not less. Hiring off-duty officers can bring genuine skill and deterrence, but it can also import constitutional-tort liability and complicate whose insurance responds (the department's, the security firm's, or yours). If a provider staffs your site with off-duty or special officers, ask specifically: under what authority do they act on my property, whose policy governs their use of force, and who indemnifies me if a § 1983 claim names me? Do not assume the clean "private citizen, no immunity" model applies.

Discrimination and civil-rights exposure in detentions

Assault and false imprisonment are not the only ways a detention goes wrong. Retail and property stops are a frequent source of discrimination and civil-rights claims — allegations that a stop was driven by race, ethnicity, or disability rather than genuine cause ("shopping while Black" cases are the well-known example). These travel under federal statutes such as 42 U.S.C. § 1981 (equal right to make and enforce contracts, which courts have applied to retail-service settings), Title II of the Civil Rights Act (public accommodations), and the Americans with Disabilities Act, as well as state and local civil-rights and public-accommodation laws. This is a distinct and severe liability channel: the damages, the reputational harm, and the discovery are different in kind from an ordinary battery claim, and a pattern of selective stops can convert individual incidents into a class or agency action. A defensible program detains on objective, articulable indicators of theft — observed concealment, not appearance — trains against profiling, and documents the specific facts that justified each stop. When you vet a provider, ask how they guard against profiling and whether their incident reports capture the objective basis for every detention.

Video is the evidence that decides use-of-force cases

Modern use-of-force litigation turns on footage. When there is clear video, it usually dictates the outcome for whichever side it favors; when there is no footage of a contested hands-on incident, the absence tends to cut against the defense — juries wonder why a professional operation didn't capture it. A written incident report is necessary but no longer sufficient. Before you sign, confirm that a provider: requires camera coverage of high-contact areas (and, where appropriate, body-worn cameras for officers likely to make contact); has a written retention policy long enough to outlast a claim's notice period; and, critically, preserves footage on any use-of-force incident so it is not overwritten before anyone requests it. Preservation-on-incident is the detail most programs get wrong, and it is the one your future defense counsel will ask about first.

De-escalation, use-of-force policy and training — what to require of a provider

The industry standard, reflected in ASIS International guidance and modern training programs, is that de-escalation is prioritized — verbal skills, creating distance, and calling for support resolve the overwhelming majority of confrontations without contact. A capable provider treats physical force as the exception that must be justified, not the default. (Current guidance, including from the Department of Homeland Security, is careful not to require officers to try lesser options before using reasonable force when facing a genuine threat; the test remains whether the officer can explain why the force used was reasonable and necessary.)

When you evaluate firms, ask to see and confirm the following:

• A written use-of-force policy that defines lawful purpose, proportionality, prohibited conduct (including a ban on chokeholds and prone restraint after control), and clear authority limits for detention, searches, and any restraint tools.
• Site-specific post orders that state when officers engage and, just as importantly, when they do not. Any weapon or tool (baton, OC spray, TASER, firearm) should be authorized only in writing and only where state law and licensing permit.
• Recurring training in de-escalation, conflict resolution, applicable state law, anti-profiling, and — where tools are carried — legal limits, medical risks, and reporting. Verify credentials against your state's rules using a license lookup, and compare programs against training requirements by state.
• Documentation and review discipline: objective incident reports capturing the articulable basis for any stop, medical aftercare offered and recorded after any force, video preserved on every use-of-force event, and prompt supervisory review (many programs target 24 hours).

Our full walkthrough on how to hire a security guard company puts these questions into a vetting checklist you can hand to any bidder.

What this means for you as the client (liability and documentation)

Two doctrines put the hiring business on the hook. Under respondeat superior / vicarious liability, the security company is generally liable for an on-duty guard's conduct within the scope of employment — and while intentional acts like assault traditionally fall outside "scope," courts widely recognize that excessive force by an on-duty guard can stay within it, because the job itself is built around the use of force. Under negligent hiring, retention, and supervision, an employer faces direct liability for its own carelessness — skipping a background check, keeping a guard with known dangerous propensities, or failing to train. That direct theory does not depend on the misconduct being within scope, and it can even reach former employees.

Property owners often assume that hiring an independent contractor shifts all of this away. It does not fully. The independent-contractor defense frequently fails for intentional torts, for conduct the owner ratified, and for non-delegable duties — and some states treat contracted guards as being under the property owner's control for liability purposes. In short, you can be pulled into a use-of-force suit even when the guard works for a firm.

Insurance is where clients get burned most quietly. Standard commercial general liability policies frequently contain an assault-and-battery (A&B) exclusion, because insurers argue an assault is an intentional act, not an "accident/occurrence." For security work — where force allegations are an occupational exposure, not a freak event — that exclusion can gut coverage for the exact lawsuit you are most likely to face. Worse, A&B coverage is often provided only as a low sublimit (for example, $25,000–$50,000) that defense costs alone can exhaust, and umbrella policies typically "follow form," so they usually will not fill an excluded or sublimited gap. Reframing the claim as "negligent hiring" often doesn't escape the exclusion either, since many are drafted broadly to reach anything "arising out of" an assault or battery.

Practical protections for you as the buyer: get named as an additional insured with verified A&B coverage; require indemnification in the contract; keep a copy of the provider's written use-of-force policy and post orders on file; and insist that incidents are reported to you promptly, documented, and backed by preserved video. Our guide to security guard contracts and insurance breaks down the exact clauses to demand.

Because use-of-force, citizen's-arrest, and defense-of-property rules vary significantly by state — as the comparison table above shows — and change over time, confirm the precise rules for your location against your state's statutes and licensing authority before you rely on them. When you are ready to compare vetted providers on these criteria — policy, training, licensing, camera coverage, and real A&B coverage — you can request quotes, browse security companies in your area, or size the budget with our security cost calculator before you talk to a single vendor.