Raine v. OpenAI Underscores Why Legacy Insurance Still Matters for AI Risk Management

Raine v. OpenAI: Why Legacy Insurance Still Matters for AI Risk Management

AI is changing how losses arise, but the liabilities look familiar. The lawsuit Raine, et al. v. OpenAI, Inc., Docket No. CGC25628528 (Cal. Super. Ct. Aug 26, 2025) makes that clear. The allegations involve a teen's death and claims that ChatGPT encouraged self-harm and provided harmful guidance. These are allegations only, and the case is at an early stage.

For insurance and risk leaders, the signal is simple: AI may be the trigger, but your first line of defense is still legacy coverage. Treat AI as a context, not a carve-out to your risk program-unless your policies make it one.

What the Case Highlights

The complaint alleges design defects, inadequate warnings, and California Unfair Competition Law violations tied to ChatGPT's responses to a disclosed intent to self-harm. Again, these are claimed facts, not proven. Regardless of the outcome, the exposure type-bodily injury-tracks long-standing liability categories.

Why Your Legacy Insurance Still Matters

CGL was built for bodily injury. If a plaintiff claims injury arising from your product, service, or operations, CGL is often where you look first. D&O responds to alleged wrongful acts by directors and officers-even where AI is part of the narrative.

As AI-related suits grow, the core lesson remains: do not overlook the policies you already buy. Often, coverage applies broadly unless an exclusion removes it.

Action Plan for Insurance and Risk Leaders

• Start with legacy coverage: Treat AI as a cause, not a coverage class. Check CGL for bodily injury and D&O for executive/board allegations. Extend the review to umbrella/excess.
• Map all possible policies: Read the complaint and any demand letters against every policy trigger. Raine-type allegations can touch CGL and D&O at once.
• Leverage other people's insurance: Vendors, suppliers, distributors, and platform partners may owe you additional insured status and indemnity. Pull contracts and request certificates and endorsements, not just promises.
• Scan for AI exclusions: Some carriers are adding AI or algorithm exclusions. If they exist, quantify the gap and address it at renewal-or via manuscript endorsements.
• Treat AI underwriting questions as high stakes: Carriers are asking about AI usage and revenue dependence. Answer precisely. Misstatements can impair coverage or support rescission.
• Engage cross-functional stakeholders: Inventory AI use across HR, operations, product, supply chain, compliance, and marketing. Different uses create different loss paths and different policy triggers.
• Review early, not after notice: Small wording differences (occurrence vs. claims-made, insured vs. additional insured, retro dates, endorsements) can swing outcomes. Involve coverage counsel before problems escalate.
• Fill true gaps with AI-specific products: If exclusions or gray areas remain, explore AI-focused solutions as a complement, not a replacement, to legacy lines.

Practical Coverage Checks to Run This Quarter

• Confirm whether "bodily injury" and "personal and advertising injury" definitions are unmodified by AI-related endorsements.
• Verify D&O definitions of "Wrongful Act," "Claim," "Insured Person," and "Securities Claim," and look for any tech/AI carve-outs.
• Audit additional insured endorsements from key partners for products-completed operations and primary/noncontributory language.
• Align incident response with notice provisions across CGL, D&O, umbrella, cyber/privacy, E&O, and media liability.
• Document AI governance and safeguards; underwriters will ask. Consistency across applications, binders, and actual practice matters.

The Bottom Line

Raine underscores a pattern: AI may be new, but the liability predicates aren't. Legacy insurance is still your foundation. Think ahead, act early, and involve the right people so exclusions, underwriting gaps, and contract misses don't turn into uninsured loss.

If your teams need practical upskilling to respond to AI-related underwriting and governance demands, explore role-based AI training resources at Complete AI Training.