Artificial intelligence is no longer a futuristic concept. It’s embedded in the car that may have rear-ended you on the 405, the medical software that generated a flawed diagnostic report, and the warehouse robot that injured a coworker. As AI systems take on more consequential roles in daily life, the question of who pays when they cause harm has become one of the most pressing issues in California personal injury law.
For years, tech companies leaned on a convenient defense: blame the algorithm. If an autonomous vehicle made a dangerous decision, or a medical AI flagged the wrong condition, developers argued the machine acted independently, and human liability was murky at best. California’s AB 316, which took effect in 2026, directly addresses that gap. It changes the legal landscape in ways that matter to anyone injured by a system with AI at its core.
Here is what you need to understand about this new law, how it interacts with existing California personal injury statutes, and what steps to take if you believe AI played a role in your injury.
What AB 316 Actually Does
AB 316 closes what had become a growing loophole in California personal injury litigation. Before this law, defendants in tech-related injury cases could argue that an AI system made an autonomous decision, effectively diffusing responsibility away from any human or corporate actor. The more sophisticated the AI, the easier it was to obscure the chain of causation.
AB 316 establishes that when a person or company deploys an AI system in a context that requires human oversight, they cannot escape liability simply by pointing to the AI’s output. If a developer or deployer knew, or reasonably should have known, that the system posed foreseeable risks and failed to address them, they remain accountable under standard negligence principles.
This is not a radical departure from California tort law. It is an application of existing negligence doctrine to a new category of defendants. California Civil Code Section 1714 has long held that everyone is responsible for injuries caused by their want of ordinary care. AB 316 clarifies that designing, deploying, or maintaining an AI system carries that same duty of care, and that duty does not evaporate because a machine made the final decision.
Where AI Injuries Are Actually Happening
The injuries we are seeing in practice are not hypothetical. They fall into a few recurring categories.
Autonomous and semi-autonomous vehicles. California has more autonomous vehicle testing miles than any other state, and consumer vehicles with advanced driver assistance systems are everywhere. When a Tesla in Autopilot mode fails to detect a stopped vehicle, or a Waymo routes a passenger into a dangerous situation, the injury is real even if no human driver was actively in control. These cases involve product liability, negligence, and now the AB 316 framework.
AI-assisted medical tools. Hospitals and clinics increasingly use AI to assist with imaging analysis, treatment recommendations, and patient triage. When those tools produce errors that delay diagnosis or lead to the wrong treatment, patients suffer. John Reardon spent 20 years as a chiropractor before becoming a lawyer, and he has seen how a missed or delayed diagnosis compounds a patient’s injury over time. Documentation of the AI tool’s role in the clinical decision becomes critical evidence in these cases.
Workplace AI systems. Automated machinery, AI-driven logistics platforms, and robotic systems in warehouses and manufacturing facilities are injuring workers. California Labor Code protections and workers’ compensation rules still apply, but AB 316 opens a path to pursue third-party claims against the developers and deployers of those systems when negligent design contributed to the harm.
AI-generated advice and reports. This category is still developing, but it includes situations where AI-generated legal, financial, or medical information causes someone to take a harmful action. The liability theory here is less settled, but the principle from AB 316, that foreseeable harm from a negligently deployed system creates accountability, is directly relevant.
Proving Negligence Against an AI Developer
AB 316 does not create strict liability. You still need to prove negligence, which means showing that the developer or deployer owed you a duty of care, breached that duty, and that the breach caused your injury and damages. What the law does is prevent defendants from using the AI’s autonomous nature as a complete shield.
In practice, proving negligence against a tech company requires evidence that most injury victims would not think to preserve on their own. The key categories include:
System logs and event data. AI systems generate detailed records of their decisions. In an autonomous vehicle crash, the vehicle’s data recorder captures sensor inputs, system states, and the decisions the AI made in the seconds before impact. This data can be overwritten or lost if not preserved quickly. We send spoliation letters to defendants early in our cases to prevent destruction of this evidence.
Developer documentation. Internal records showing what risks the developer identified during testing, what safety measures were considered and rejected, and what the system’s known failure modes are can establish that the harm was foreseeable. This material is often obtained through discovery, which is another reason to retain counsel before the evidence trail goes cold.
Regulatory filings. California’s Department of Motor Vehicles requires autonomous vehicle developers to report disengagements and accidents. The National Highway Traffic Safety Administration maintains similar records at the federal level. These public filings can corroborate your account of what happened and show a pattern of similar failures.
Expert testimony. AI liability cases almost always require expert witnesses who can explain to a jury how the system worked, what it should have done differently, and why the developer’s choices fell below the applicable standard of care. Identifying and retaining the right experts early is essential.
How Insurance Companies Are Responding
Insurance carriers have not been passive observers of this legal shift. We are already seeing adjusters and defense teams deploy specific tactics in AI-related injury claims.
The most common is minimizing the AI’s role entirely. An adjuster handling an autonomous vehicle claim may try to reframe the incident as a conventional car accident, focusing on road conditions or the other driver’s behavior rather than the system’s failure. This framing keeps the claim within familiar parameters and away from the larger liability exposure that AB 316 creates.
A second tactic is delay. AI liability cases are complex, and defendants know that complexity favors the party with more resources. The longer a case drags on, the more pressure an injured person feels to accept a lower settlement. We push back on delay by moving aggressively in discovery and, where appropriate, by filing suit rather than waiting for a negotiated resolution that never materializes.
A third approach is to argue that the injured person assumed the risk by using an AI-enabled product. This argument has limited traction under California’s comparative fault framework, but it can affect the damages calculation if a jury finds partial fault. Thorough documentation of how the product was marketed and what safety representations the developer made is the best counter to this defense.
What to Do If AI Was Involved in Your Injury
The steps you take in the days and weeks after an AI-related injury can determine whether you have a viable claim.
First, document the AI’s involvement as specifically as possible. If you were in a crash involving a vehicle with autonomous or semi-autonomous features, note the make, model, and any information about what driving mode was active. If a medical AI tool was involved in your care, request your complete medical records, including any documentation of the software used in your diagnosis or treatment.
Second, do not interact with the at-fault party’s insurance company without counsel. This is true in any personal injury case, but it is especially important here. Statements you make about the technology involved can be used to shift blame or minimize the developer’s role.
Third, contact an attorney promptly. California’s general statute of limitations for personal injury claims is two years from the date of injury under Code of Civil Procedure Section 335.1. But evidence in AI cases, particularly system logs and internal records, can disappear well before that deadline if no one takes steps to preserve it. The sooner we are involved, the better positioned we are to build a complete record.
The Bigger Picture for California Injury Victims
AB 316 represents a meaningful shift in how California law treats the relationship between technology and accountability. For too long, the complexity of AI systems served as a practical barrier to recovery for injured people. Developers could point to the algorithm, the algorithm could not be sued, and victims were left without a clear path to compensation.
That dynamic is changing. California has signaled, through this legislation and through the broader trajectory of its consumer protection and product liability law, that deploying powerful technology in high-stakes environments carries real legal responsibility. That is good news for injured Californians, but only if they understand their rights and act on them.
We work with clients across a range of AI-related injury scenarios, from autonomous vehicle crashes to medical device failures. John Reardon’s clinical background gives us a practical understanding of how these injuries develop and what documentation is needed to support a strong claim. If you believe AI played a role in your injury, we are glad to review your situation at no cost and help you understand whether AB 316 and California’s existing personal injury framework give you a path forward.
Contact Reardon Injury Law to schedule a free consultation. There is no fee unless we recover compensation for you.