Proving AI Was Wrong: The Challenge of Modern Medical Negligence
In the past twenty years, we have witnessed some of the most significant technological advancements, including notable progress in artificial intelligence. This progress, while exciting, has introduced new challenges, especially in the legal field.
It may be hard for patients harmed by AI-assisted diagnostics, treatment suggestions, or robotic systems to prove that technology played a role in their medical negligence claim. Ohio medical malpractice lawyers have their work cut out for them in proving AI was wrong. Below, we discuss possible issues caused by AI in these cases, how these difficulties relate to traditional medical malpractice law, and what to consider when navigating this new area of liability and patient rights.
Why AI errors complicate medical malpractice claims
AI adds a layer of complexity to traditional frameworks for proving medical malpractice. Determining who’s legally responsible when humans have been influenced or directed by algorithms and automated systems is complex.
AI as a tool, not a decision-maker
AI is not the end-all and be-all, and courts agree. AI is not a licensed healthcare professional, so it can’t substitute medical advice. Courts generally treat AI as a tool or medical device. As such, when AI makes a mistake, the law often holds the doctor or facility that used, or failed to adequately question the AI, accountable. If negligence is detected, AI itself isn’t found “negligent”; but rather the party who unreasonably relied on it.
Explainability and black box systems
Many, if not most, AI models are obscure, which speaks to the commonly used “black box” reference. Sometimes, developers even struggle to explain how a particular result occurred clearly. In the legal context, this opacity makes it hard, if not impossible, to trace the chain of events from decision to injury. The ability of a system to explain its reasoning is a turning point in the allocation of responsibility between software and tech developers and medical providers.
Who’s really at fault?
Several potential issues arise with AI, including design flaws, algorithmic bias, deficiencies in data training, or misuse by clinicians. Because of the complex and distributed nature of AI, injured victims may name multiple parties in their lawsuit, including, but not limited to, device manufacturers, data vendors, hospitals and care facilities, and individual practitioners. Untangling this web of multiple actors is a substantial reason why proving direct negligence is burdensome.
How traditional Ohio medical malpractice law frames AI’s mistakes
Although Ohio law does not yet directly address AI-specific medical negligence, there are some general standards that still apply.
Four elements of a medical malpractice claim
In general, to prove medical negligence, the plaintiff must prove four elements:
- Duty owed by the healthcare provider to the patient
- Breach of the standard of care
- Causation (the breach caused harm)
- Damages (compensation for loss or injury)
The plaintiff’s claim must also include an Affidavit of Merit (AOM), which is a sworn statement from a medical expert. The AOM acknowledges that the claim has a reasonable basis as the provider’s conduct fell below the appropriate standard.
Loss of chance doctrine
Under this doctrine, a previous patient can receive compensation if they had a diminished chance of recovery due to a doctor’s negligence. Even in cases where recovery was relatively unlikely, plaintiffs are still entitled to damages (damages are awarded proportionally to the lost chance, not for the entire injury).
Suppose an AI error is the foundation of a claim, an Ohio medical malpractice lawyer might argue that improper interpretation or reliance on AI negatively impacted the ability to offer timely diagnosis or treatment. Of course, this argument must be rooted in expert evidence.
Statute of limitations and statute of repose
Those harmed by medical negligence have two statutes to consider when deciding whether to file a claim. Ohio has a 1-year statute of limitations for medical malpractice actions that starts from the date the injury is or should have been discovered.
Additionally, there is a 4-year statute of repose. According to ORC § 2305.113(c), except for certain exceptions, a claim can’t be brought more than four years past the negligent act, regardless of late discovery. Per the Ohio Supreme Court, the statute of repose also applies to wrongful death claims due to medical error (wrongful death claims may be barred when based on time-barred medical claims).
Unique challenges of proving AI error in Ohio
Proving medical negligence becomes even more nuanced when AI is in the mix.
Requirement of expert testimony
AI and medical decision-making alike are incredibly technical. As such, courts require expert witnesses who can explain:
- The standard of medical care
- How AI impacts that standard
- How the practitioner uses or overrides (or fails to) deviate from that standard
- How the deviation caused harm
Identifying experts versed in both medicine and AI systems is harder, especially since many clinicians lack AI training.
Proving causation through AI paths
Not unlike standard negligence claims, an Ohio medical malpractice lawyer must prove that but for the provider’s breach, the plaintiff wouldn’t have sustained the injury. For cases involving AI, the link between a medical decision and harm can be difficult to discern. Plaintiffs are faced with a chain of causation in a multi-step environment. Credible data, algorithms, or engineering audits of AI systems help make the correlation.
What if you can’t prove AI was wrong?
In exceptional cases where it’s incredibly challenging to prove causation, a plaintiff may invoke the doctrine of res ipsa loquitur. This Latin phrase translates to “the thing speaks for itself.” Here, courts will allow negligence to be inferred if the harm wouldn’t ordinarily occur without negligence in play – although this is rarely permitted in medical malpractice claims.
In AI-related medical malpractice cases, plaintiffs might invoke this doctrine when proprietary code and hidden processes obstruct discovery. Due to the limited applicability of this doctrine, courts will scrutinize whether the prerequisites for res ipsa loquitur are met.
Multiple responsible parties and shifting liability
AI, as the basis of medical negligence, may not only implicate the doctor and/or healthcare provider but also possibly the software vendor. In cases involving multiple defendants, the proceedings may be bifurcated, splitting into two distinct parts: product liability and medical negligence.
The product liability aspect may focus on a defect in the device, and the medical negligence aspect could concentrate on misuse or overreliance. An Ohio medical malpractice lawyer can manage the side effects of the defense’s strategies and crossclaims, like delays, motion practice, and disputes over liability allocation.
What happens when a doctor admits fault?
In medical malpractice cases, Ohio law considers expressions of sympathy inadmissible, but explicit admissions of negligence are not protected. So, if a medical provider acknowledges that AI made a mistake, that statement admitting fault may be admissible depending on content and context.
Example: AI misreads imaging and fails to detect a stroke
A radiology AI system flags a brain MRI as normal. Even though the radiologist notices a seemingly minor lesion, they assume the AI is accurate and don’t verify the imaging. Hours after the MRI, the patient had a stroke that could’ve been prevented by early intervention.
Following the four-part negligence test:
- The plaintiff could argue the radiologist owed them a duty of care and breached that duty by failing to review the imaging.
- An expert could explain how a reasonable radiologist would’ve verified the negative AI output, especially when clinical signs were present.
- The plaintiff’s attorney can make the case that if the lesion had been caught, treatment could’ve been provided to reduce damage.
- Consequently, the patient sustained greater disability than would have otherwise occurred.
Injured by an AI error? Act before time runs out
Due to the state’s strict statutes of limitations and repose, if you or someone you know has been injured due to a medical AI error, you should act quickly. Though proving AI was wrong presents new and unfamiliar hurdles, there are still paths to accountability, and no need to go it alone.
Contact us via phone or online to schedule a consultation. Crandall & Pera Law’s team of Ohio medical malpractice lawyers works closely with experts in medicine and technology to build compelling cases.