Duty of care is the first question in any negligence case, and the one that decides whether the case ever reaches a jury on the other elements. Before a court asks whether a defendant breached a standard, caused an injury, or owes damages, it asks a narrower question: did this defendant owe this plaintiff any legal obligation to be careful in the first place? If the answer is no, the case ends there, no matter how badly the plaintiff was hurt or how careless the defendant's conduct looked.

Duty is a question of law decided by the judge, not a question of fact decided by the jury. That single feature makes it the doctrine most likely to end a negligence claim at the pleading or summary-judgment stage, long before trial. Litigators fight over duty precisely because it is the cheapest place to win or lose the whole case.

Most jurisdictions start from the foreseeability principle: a duty of care generally runs to anyone within the foreseeable zone of danger created by the defendant's conduct. If a reasonable person in the defendant's position should have foreseen that careless conduct could injure the plaintiff or someone in the plaintiff's position, a duty typically exists.

Some jurisdictions supplement or replace pure foreseeability with a multi-factor balancing test drawn from the Restatement (Third) of Torts and similar sources -- weighing foreseeability against the burden of precautions, the closeness of the connection between conduct and injury, moral blame, the policy of preventing future harm, and the availability of insurance. Courts diverge on how much weight foreseeability alone should carry versus these policy factors, and that divergence is itself often the crux of a duty fight.

The common-law baseline is that a person owes no affirmative duty to rescue or protect a stranger from harm the person did not create. This is one of the most counterintuitive features of negligence law to lay juries and clients alike -- a bystander who watches a drowning and does nothing is typically not liable, absent more.

Courts carve out well-recognized exceptions where a special relationship creates an affirmative duty to protect or warn: common carrier and passenger, innkeeper and guest, employer and employee, school and student, landlord and tenant in some jurisdictions, and custodian and dependent. A defendant who voluntarily undertakes to help can also create a duty it did not otherwise have, if the undertaking increases the risk of harm or the plaintiff reasonably relies on it.

Because duty is decided by the judge as a matter of law, it is where courts do their most openly policy-driven work in negligence -- deciding, in effect, how far liability should extend before litigation becomes unmanageable or deters socially useful conduct. Landowner duties (owed differently to invitees, licensees, and trespassers in states that retain the classifications), professional duties defined by a professional standard of care, and duties limited by the economic loss rule are all examples of duty being narrowed or shaped for policy reasons rather than pure foreseeability.

This is also where courts most often decline to extend liability even when harm was technically foreseeable, citing the risk of limitless liability to an indeterminate class of plaintiffs.

In Juricratic, duty is modeled as the gating decision node at the root of a negligence claim path -- before breach, causation, and damages dials ever activate. Because a no-duty ruling ends the case as a matter of law, it is the highest-leverage node to war-game early: a plaintiff-side simulation should test how sensitive the outcome distribution is to the facts a court would use to find (or reject) a special relationship or foreseeable zone of danger, while a defense-side simulation can model the payoff of an early dispositive motion attacking duty against the cost and delay of litigating breach and causation instead.

Sweeping the duty dial from clearly-owed to clearly-not-owed shows counsel exactly how much of the claim's expected value depends on a threshold ruling that a judge, not a jury, will make.