Google's AI Overview Invented a Lawsuit That Destroyed a Solar Company's Sales
What happened
When potential customers searched for Wolf River Electric online, what they found was a Google AI-generated summary telling them the Minnesota solar company was under investigation by the state Attorney General for misleading consumers and using high-pressure sales tactics. None of it was true. The company had never been sued by the Attorney General. But the contracts started canceling anyway.
The AI Overview system cited four sources for its claim, including a Star Tribune article, an Angie's List listing, a press release from the Minnesota Attorney General's office, and a KROC News story. According to Wolf River's complaint, not one of those sources actually states that the company was sued. What the system appears to have done is combine details from unrelated legal actions against other solar companies and attach them to Wolf River's name. The result was an authoritative-looking summary about a lawsuit that did not exist.
The business damage was concrete. Customers canceled projects when they saw the summaries. Wolf River says it lost over USD 25 million in sales during 2024, with specific canceled contracts totaling more than USD 150,000 traced directly to the false information. The company raised the errors with Google. Google, according to the complaint, offered no correction, no retraction, and no public acknowledgment that the summaries were wrong. The lawsuit, seeking over USD 110 million in damages, followed.
This incident sits in a narrower category than most AI misinformation cases. The system did not fabricate a vague negative impression. It fabricated a specific legal event, attributed it to a named business, and cited real sources to support the claim, sources that did not actually say what the summary claimed they said. That is a defamation mechanism, not a rounding error, and it ran at search scale on a platform people use as a first stop when vetting a contractor they are about to hire.
The harder problem is not the hallucination itself but the absence of any correction path once the error was reported. A system that generates a legal claim about a business and then surfaces that claim to hundreds of prospective customers leaves no retrievable record of what sources were combined or how, and no clear mechanism for the business to force a retraction. A provable record of what a system retrieved, how it synthesized a claim, and whether that claim was ever challenged and corrected would not have prevented the initial error. It would have made the error addressable rather than permanent.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.