AI Evidence in Ohio Malpractice Cases: Risks & Opportunities

Artificial intelligence, AI for short, is transforming life as we know it. AI has had a particularly significant impact in the healthcare industry, from diagnostic imaging to predictive analytics. However, the results of AI tools haven’t been easily integrated into other fields, such as law, for example. In particular, AI as a component of medical malpractice lawsuits in Ohio has been a learning process for litigants.

Given the numerous considerations involved in navigating this new frontier of risk and opportunity, we’ve decided to take a closer look at recent developments. In this article, we examine how AI-based material may be used or challenged in malpractice cases, identify legal pitfalls to avoid, and outline strategies employed by Ohio medical malpractice lawyers to help preserve and scrutinize AI evidence.

The rising role of AI in medicine

Medical providers increasingly rely on AI models. Most providers aren’t using AI because they lack the necessary competence or experience. They’re likely very qualified professionals who appreciate technological tools that help them perform more effectively and efficiently. Common AI systems used for medical tasks include:

• Interpreting radiology or pathology images
• Predictive prognostics
• Risk stratification
• Decision support, like recommending treatment options
• Automating documents
• Summarizing patient records

When AI makes any type of contribution to clinical decision-making or record generation, the results may surface amid litigation. The plaintiff or the defense can use these results as evidence. Herein lies the problem with AI as evidence. Often, AI models are opaque, frequently referred to as “black boxes.” These black boxes raise unique questions regarding reliability, explainability, and authentication.

Key legal issues: Admissibility, authentication, and expert testimony

It’s not always clear how to apply AI evidence in medical malpractice lawsuits. There are some prominent legal hurdles to overcome.

Admissibility under Ohio evidence rules and Daubert standards

The Ohio Rules of Evidence Rule 702 require that the proponent show by a preponderance of the evidence that expert testimony is based on reliable scientific, technical, or specialized information and that the expert reliably applied the methods to the facts. Although general acceptance may be considered, it is not required.

Proposed FRE 707 would require machine‑generated evidence to satisfy Rule 702’s reliability requirements (particularly when offered without a human expert), but it does not make AI an ‘expert witness.’ Although this rule has been proposed but not yet adopted, it does signal the direction in which the law is likely to head.

Authentication of AI evidence

Validity is just one hurdle that AI output must overcome. While the production may very well be valid, it must also be authenticated under Ohio’s Evidence Rule 901. According to this rule, evidence must be “what the proponent claims it is.” Currently, AI evidence isn’t directly addressed under this rule. However, newly proposed federal amendments to Rule 901 would insert explicit guidance if adopted.

In real life, AI outputs are typically challenged by litigants contesting:

• Who developed or trained the AI system
• Whether the version in question was modified or altered in any way
• Whether the input data was biased or corrupted
• Whether the output is the original, unaltered result

From these common inquiries, it’s no wonder that the chain of custody of AI evidence is crucial. Consider a hospital deleting logs or failing to maintain version history; it may become impossible to validate the output.

Cross-examination and black box problems

As mentioned earlier, AI models aren’t entirely transparent. Clinicians or other healthcare professionals may not be aware of or understand how inputs are mapped to outputs. During cross-examinations of experts, opposing parties may exploit that vulnerability by questioning the system’s reasoning or hidden biases. If the AI model’s logic can’t be adequately explained straightforwardly, the court will probably reject the AI output as speculative or unreliable.

Risks to plaintiffs and defendants in using AI evidence

In Ohio, AI evidence may offer compelling insights or support complex claims, but all parties to a lawsuit face risks in relying on AI-derived materials.

Potential challenges to plaintiffs

• Without timely preservation of version history, input logs, and metadata, critical output can be overwritten.
• To ensure that AI systems function effectively, many hospitals and healthcare facilities regularly upgrade them. So, the version used at the time of care might not be the same as the one used throughout litigation.
• The defense may hire its own data scientist or AI expert to question the model’s legitimacy in terms of training, input data, output data, or bias.
• AI is still relatively new to most people, so some judges and juries may view AI evidence as untrustworthy.

Pitfalls defendants may encounter

• Healthcare providers and institutions that rely too heavily on AI systems without sufficient human oversight may expose themselves to negligence claims. The plaintiff can argue that they heavily depended on unvalidated technology.
• The plaintiff is also able to seek disclosure of training data, model weights, or even source code. Providing this type of information could compromise confidentiality, trade secrets, or proprietary information, as such discovery is often contested and subject to protective orders or denial.
• Depending on the defendant’s disclosures, plaintiffs may argue that the AI model systematically hurt their claims by being biased against particular demographics.

Strategies for leveraging and defending AI evidence

Using materials that come from AI requires careful planning and technical know‑how.

Early preservation and notice

At the outset of a case, legal counsel should immediately issue a preservation letter demanding that the institution or provider preserve certain information. Specifically, the letter can demand a freeze placed on AI logs, model versions, training data, metadata, and system documentation to prevent tampering or deletion.

Additionally, it’s a good idea to notify opposing counsel and the court about your intent to use AI evidence. If disagreements occur over who can access evidence or whether evidence is admissible, they can be settled sooner rather than later.

Expert support and validation

Ohio medical malpractice lawyers can leverage AI evidence more effectively by collaborating with experts in the field, such as data scientists or machine learning engineers. These experts can assist by:

• Validating the model’s accuracy, error rates, and restrictions;
• Reproducing results on mirrored data;
• Explaining black box logic in simple terms, and
• Simulating different inputs and conducting sensitivity analyses.

On the other hand, defense counsel can hire counter-experts to assist in planning counterattacks or suggest alternative AI systems.

Motion practice: Motions in limine, Daubert motions

Before the start of the trial, either party can file motions to exclude AI evidence. At this point, judges will evaluate the strength and reliability of the science. If the motion is successful, the evidence could be thrown out, which in turn could significantly affect the outcome of the case.

Deep dive into AI systems

Counsel should utilize legal instruments, including interrogatories, depositions, and requests for document production, to understand how the model operates. Key areas of interest may cover:

• System architecture, training data, version logs;
• Performance metrics and validation studies;
• Policies on maintenance, updates, and governance; and
• Reports are issued throughout system development.

Discovery may reveal weaknesses or prejudices that are not readily apparent.

Jury education

If motions to exclude AI evidence don’t work, Ohio medical malpractice lawyers should take advantage of visual aids, basic analogies, and credible expert testimony to teach jurors. The jury should understand how AI works, its limitations, and how much weight to assign it.

Balancing innovation of AI evidence and legal rigor in Ohio malpractice cases

AI evidence is a monumental tool, and also a potential minefield in Ohio. Plaintiffs and defendants alike must be strategic in how they use this kind of evidence. As courts learn and adapt, the most effective litigators will know how to balance innovation with the evolution of law and public opinion.

If you or someone you love has a medical malpractice claim involving AI, please reach out to us today. Our Ohio medical malpractice lawyers at Crandall & Pera Law know how to investigate complex technical evidence and make strong arguments in court.