Ohio Supreme Court Restricts Discovery of Insurer Claim Files in Bad-Faith Cases

When a policyholder sues an insurer for bad faith, the discovery fight routinely overshadows the fight over the merits. Plaintiffs’ counsel want the claim file, the adjuster’s notes, the reserve memos, the coverage opinions, and the communications with defense counsel, because it is usually the best evidence of how the insurer actually evaluated the claim, as opposed to how the insurer now says it did. Insurers, for the same reason, treat that file as their most sensitive litigation asset and resist producing anything that touches legal advice or claims strategy. On February 26, 2026, the Supreme Court of Ohio waded squarely into that fight. In Eddy v. Farmers Property Casualty Insurance Co., the court held that Ohio’s attorney-client privilege statute, not a judge-made bad-faith exception, controls whether an insurer’s privileged claim-file communications must be disclosed, and that work-product materials in the file remain separately protected absent an independent showing of good cause.¹ The decision overturns two decades of lower-court practice built on the court’s own 2001 decision in Boone v. Vanliner Insurance Co.², and it gives Ohio insurers a considerably stronger hand in resisting broad claim-file discovery requests. For bad-faith litigators outside Ohio, Eddy is also a useful data point in an unsettled, closely watched area of law: how far can a policyholder go in prying open an insurer’s privileged deliberations simply by alleging that the insurer acted unreasonably?

Why the Claim File Is the Central Battleground

Claim-file discovery disputes recur in bad-faith litigation nationally for a structural reason. The tort of bad faith typically turns on the insurer’s state of mind, whether it had a reasonable basis for denying, delaying, or underpaying a claim, and whether it knew of or recklessly disregarded the unreasonableness of its position. Direct evidence of that internal reasoning lives almost exclusively in the claim file: adjuster diaries, reserve worksheets, underwriting and coverage analyses, and communications with in-house or outside counsel about how to handle the claim. A policyholder who cannot see that file is left arguing about the insurer’s state of mind from the outside, using only the letters and payments the insurer chose to send. An insurer forced to produce the file without limitation, on the other hand, loses the candid, privileged legal analysis that the attorney-client privilege and work-product doctrine are designed to protect, and that insurers argue is essential to sound claims handling in the first place. Courts have wrestled with that line for decades and have not landed in the same place.

Three Competing Approaches Nationally

Several doctrinal frameworks have emerged. A handful of states apply a version of the fiduciary exception, treating an insurer’s relationship to its policyholder as sufficiently trust-like that ordinary privilege protections give way once the insured alleges the insurer breached that duty. Washington has gone furthest: in Cedell v. Farmers Insurance Co. of Washington, the state supreme court held that in first-party bad-faith cases other than UIM claims there is a presumption of no attorney-client privilege over the insurer’s claim-handling communications, rebuttable only by an in camera showing that counsel was providing pure coverage advice rather than performing a quasi-fiduciary claims function.³ Other states rely on implied-waiver doctrine instead of a bad-faith-specific exception. The Arizona Supreme Court held in State Farm Mutual Automobile Insurance Co. v. Lee that an insurer waives the privilege only when it affirmatively places counsel’s advice at issue, such as by defending a bad-faith claim on the ground that its claims personnel subjectively understood the law to permit its coverage position, where that understanding was formed in part on advice of counsel. Unlike Cedell’s presumption of no privilege, Lee presumes the privilege applies and asks whether the insurer’s own litigation conduct forfeited it; an insurer that simply denies bad faith and defends on objective grounds does not waive.⁴ Mississippi courts have taken a related but distinct approach, focused on whether the claims handler actually relied on counsel’s advice in denying or defending the claim. The Mississippi Supreme Court held in Travelers Property Casualty Co. of America v. 100 Renaissance, LLC that an insurer waived privilege over its in-house lawyer’s advice where the claims adjuster’s denial letter tracked the lawyer’s reasoning so closely that the adjuster could not explain the letter’s own legal basis at her deposition, even though the insurer never formally asserted an advice-of-counsel defense.⁵ Still other jurisdictions treat the claim file as generally discoverable evidence of the insurer’s claims-handling process, while recognizing that privilege can attach to particular communications and that work product remains protected absent a substantial-need showing. Ohio’s own pre-Eddy caselaw belonged in that last camp, and arguably went further than most.

Ohio’s Path to Eddy: Moskovitz, Boone, and the 2007 Statutory Fix

The Boone line traces to a 1994 decision, Moskovitz v. Mount Sinai Medical Center, in which the Ohio Supreme Court held that claim-file materials showing an insurer’s lack of a good-faith effort to settle a case were unworthy of privilege protection in a narrow, postjudgment proceeding for prejudgment interest.⁶ In 2001, Boone extended that reasoning well beyond the prejudgment-interest context, holding that in any action alleging bad-faith denial of coverage, an insured could obtain claim-file materials containing attorney-client communications related to coverage that were created before the insurer denied the claim, on the theory that such materials could not yet contain protected work product because coverage had not yet been decided.⁷ The decision drew a pointed dissent warning that a mere allegation of bad faith should not be enough to strip an insurer of privilege, and it was criticized in leading insurance treatises as an outlier nationally. The Ohio General Assembly responded in 2007, amending the privilege statute, R.C. 2317.02(A)(2), to provide that an insurer’s attorney may be compelled to testify about attorney-client communications only if the party seeking disclosure first makes a prima facie showing of bad faith, fraud, or criminal misconduct, and even then only as to communications that a court, after an in camera inspection, determines relate to the attorney’s aiding or furthering an ongoing or future commission of bad faith.⁸ The legislature also adopted uncodified language stating expressly that the amendment was intended to modify the common law established in Boone and Moskovitz. Despite that language, many Ohio appellate courts continued to apply Boone’s broader rule in bad-faith discovery disputes while others read the statute narrowly, leaving its actual reach unresolved until Eddy.

The Eddy Litigation

Eddy arose from an ordinary underinsured-motorist dispute. Melissa and Alexis Eddy were injured in a 2020 car accident, recovered the $100,000 limit of the at-fault driver’s policy, and then sought the difference under their own UIM policy with Farmers Property Casualty Insurance Co. Farmers offered $33,312 against the Eddys’ $150,000 demand, and the parties litigated the coverage dispute for months before Farmers ultimately paid the full $150,000 after receiving updated medical information. The Eddys then filed a second suit alleging that Farmers had acted in bad faith by delaying resolution of the claim, and in discovery they sought Farmers’ complete claim file, including materials created after they had already sued Farmers over coverage. Farmers withheld twenty documents from that period as privileged or protected work product and produced a privilege log describing each one, but the trial court ordered the entire file produced without conducting the in camera review that R.C. 2317.02(A)(2) requires. The First District Court of Appeals affirmed, reasoning that the Eddys’ bad-faith allegation, combined with what it viewed as an insufficiently detailed privilege log, was itself enough to justify disclosure under Boone.⁹

What the Ohio Supreme Court Held

The Ohio Supreme Court reversed, 5–2, in an opinion by Justice DeWine; the two dissenters would have dismissed the appeal as improvidently accepted rather than reach the merits. The court held that the 2007 amendment to R.C. 2317.02(A)(2) completely abrogated Boone as it applies to the attorney-client privilege, because the statute provides a comprehensive framework, a prima facie bad-faith showing followed by an in camera review limited to communications that further an ongoing or future commission of bad faith, that leaves no room for a broader common-law exception.¹⁰ The opinion is notable for treating the attorney-client privilege and the work-product doctrine as governed by entirely separate rules, a distinction some lower courts had blurred. On the privilege side, the court rejected the argument that R.C. 2317.02(A)(2) applies only to live testimony and not to documents, relying on a 1936 decision holding that a statute barring an attorney from testifying about client communications also bars discovery of documents that reveal those communications.¹¹ Because the statute’s own text requires an in camera inspection, a term the court noted ordinarily describes a judge’s private review of documentary evidence, and because the legislature’s uncodified findings referred broadly to all communications between an attorney and a client, the court concluded the statute governs claim-file documents as much as deposition or trial testimony.

On the work-product side, the court held that Boone never displaced Civil Rule 26(B)(4) in the first place. Boone’s own syllabus was expressly limited to attorney-client privilege, and the Boone court had declined to order disclosure of anything it considered work product. Work-product materials in an insurer’s claim file are therefore presumptively protected and discoverable only on the traditional good-cause showing, that the material is directly at issue, the need for it is compelling, and the evidence cannot be obtained elsewhere.¹² Applying both standards, the court remanded the case with explicit instructions: the trial court must first decide whether the Eddys made a prima facie showing of bad faith, fraud, or criminal misconduct; if so, it must review the withheld documents in camera and disclose only those attorney-client communications that relate to furthering bad faith; and, separately, it must determine whether good cause supports disclosing any work product under Civil Rule 26(B)(4), a determination the court indicated will often require an in camera review in cases like this one.¹³ The court also rejected the appellate court’s alternative holding that Farmers had forfeited any claim of privilege by failing to describe the withheld documents in sufficient detail, finding that Farmers’ privilege log and its repeated requests for an in camera review were adequate, and that the Eddys had never raised forfeiture themselves.

Practical Implications for Insurers and Policyholder Counsel

First, Ohio insurers should treat the privilege log as a document that will be scrutinized, not a formality. Eddy makes clear that a sufficiently detailed log, coupled with a timely, repeated request for in camera review, will generally satisfy an insurer’s initial burden under Civil Rule 26(B)(8) and preserve the two-step statutory analysis the court has now mandated. Insurers that produce vague or boilerplate privilege logs, or that fail to press for in camera review at the trial court level, risk forfeiting the very protections Eddy otherwise affords them.

Second, policyholder counsel litigating bad-faith claims in Ohio can no longer treat a bad-faith allegation, by itself, as a key that unlocks the claim file. Counsel will need to develop an evidentiary record, often through the portions of the file already produced, through correspondence, or through early depositions of claims personnel, sufficient to make a prima facie showing of bad faith, fraud, or criminal misconduct before a court will even reach the question of what privileged material should be reviewed in camera. Motions to compel that rest solely on the existence of a bad-faith complaint are unlikely to succeed after Eddy.

Third, both sides should expect the attorney-client privilege and work-product analyses to proceed on separate tracks, with separate standards, going forward. A document that survives the statutory prima facie and in camera review as an attorney-client communication may still be independently analyzed as work product, and vice versa. Litigants and courts will need to categorize each withheld document carefully rather than treating the claim file as a single undifferentiated pool subject to one bad-faith exception.

Fourth, Eddy’s persuasive value outside Ohio will depend heavily on how a given jurisdiction has already structured its own privilege doctrine. States that, like Ohio before 2007, rely on judge-made bad-faith exceptions without a governing statute may find Eddy’s insistence on a threshold showing and in camera review a useful model for reining in overbroad claim-file discovery, particularly where legislatures have shown any interest in codifying insurer privilege protections. But jurisdictions that have already adopted a fiduciary-exception presumption, as Washington has in Cedell for non-UIM claims, or that rely on implied-waiver doctrines built around an insurer’s own litigation conduct, as Arizona does in Lee, and as Mississippi does through its reliance-based waiver analysis in 100 Renaissance, are unlikely to abandon those frameworks simply because Ohio has taken a different statutory path. Eddy is best read as persuasive authority for the proposition that bad-faith allegations alone should not defeat privilege, not as a template that will displace well-established doctrine elsewhere.

Fifth, for practitioners handling bad-faith claims across multiple jurisdictions, including Arizona and Mississippi, where much of Ohio’s underlying doctrinal debate has direct analogues, the safest course remains to manage privilege risk at the claims-handling stage rather than relitigate it after suit is filed. Claims personnel should be trained to document the factual and business basis for coverage decisions independently of legal advice wherever possible, so that an insurer defending a bad-faith claim is not forced to choose between waiving privilege and being unable to explain its own reasoning, the trap that undid the insurer’s privilege claim in Mississippi’s 100 Renaissance and that triggers an implied waiver under Arizona’s Lee standard. Coverage counsel should also be looped in early enough, and with a clear enough division between business claims-handling advice and litigation-driven legal analysis, that a later privilege fight turns on a well-organized file rather than a reconstructed one.

Looking Ahead

Eddy does not end claim-file discovery disputes in Ohio bad-faith litigation; it channels them. Trial courts must now conduct the two-track analysis the Supreme Court laid out on remand in Eddy itself, and appellate courts will likely spend the next several years working out how demanding the prima facie showing must be and what kinds of documents satisfy the good-cause standard for work product. For insurers, the decision restores meaningful protection to claims-handling communications that many had assumed, under Boone, were essentially unprotected once a bad-faith complaint was filed. For policyholder counsel, it raises the bar for obtaining that material but does not eliminate the path to it. And for bad-faith litigators watching from Arizona, Mississippi, and other jurisdictions still working out their own rules in this area, Eddy is a reminder that the national split over claim-file discovery is unlikely to resolve into a single uniform rule anytime soon, which makes jurisdiction-specific privilege strategy, not a one-size-fits-all playbook, essential to defending or prosecuting a bad-faith case.