National survey of 500 employed AI users: 42% have knowingly accepted an answer they suspected was wrong, 30% have already had a problem at work because of one, and only 22% work under a written rule requiring them to verify AI output.

IRVINE, Calif., Sept. 2, 2026 /PRNewswire/ — Most American workers who use AI on the job already understand who pays when a wrong AI answer causes harm: they do, not the company that built the tool. That knowledge has not changed how they work. A national survey released today by Kolmogorov Law, P.C., a California business litigation firm, finds that 65 percent of working AI users do not always verify an AI answer before acting on it or passing it along. Forty-two percent admit they have gone ahead with an answer they suspected was wrong.

The survey of 500 employed U.S. adults who use AI tools for work was fielded September 1, 2026 through the Pollfish research platform. It measured how often workers check AI output, what happens when they don’t, and who they believe is legally responsible when it goes wrong.

Asked who would be legally responsible if they acted on a wrong AI answer at work and it caused financial harm, 51 percent of respondents said “me personally” and 18 percent said their employer. Only 12 percent named the company that made the AI.

“Workers have figured out something most of their employers have not written down yet,” said Pavel Kolmogorov, founding attorney of Kolmogorov Law. “Half of them know the liability for a wrong AI answer is theirs. They are still not checking. When that answer ends up in a contract, a filing, or a client deliverable, ‘the AI said so’ is not a defense anyone has won with.”

The survey found that the checking gap is already producing consequences. Thirty percent of respondents said a wrong or inaccurate AI answer has caused a problem in their work. The problems ranged from errors in deliverables and bad decisions to lost time or money and client or legal issues. Among the 113 respondents who use AI for legal, financial, or compliance questions, that figure rose to 43 percent.

Employers have largely left the question unaddressed. Only 22 percent of respondents said their employer has a written policy requiring AI output to be verified before it goes into work product. Fifty-nine percent said no such policy exists and 19 percent did not know.

Many workers are also unaware of the fine print. Fifty-two percent did not know that the terms of service for major AI products disclaim responsibility for the accuracy of their answers. In practice, that leaves users without a claim against the vendor.

“The AI companies have disclaimed accuracy in their terms of service, so there is no one upstream to sue,” Kolmogorov said. “That leaves the professional and the business, which is exactly where the duty of care has always been. The tool changed. The rule did not.”

On one point respondents were nearly unanimous. Seventy-eight percent said professionals such as lawyers, accountants, advisors, and doctors should be required to disclose to clients when AI was used in their work. Among respondents who are themselves client-serving professionals, support reached 84 percent.

The findings arrive as courts begin to address AI accuracy directly. On May 28, 2026, the Regional Court of Munich held Google liable for false statements generated by its AI Overviews, rejecting the argument that instructing users to double-check the results was sufficient. Google is appealing.

The full survey results, methodology, and analysis are available at https://www.kolmogorovlaw.com/cognitive-surrender-65-percent-of-workers-dont-always-check-ai-answers.

Key findings

• 65% of workers who use AI on the job do not always verify an AI answer before acting on it; 32% verify only sometimes, rarely, or never.
• 42% have knowingly accepted an AI answer they suspected was wrong because it was faster or sounded confident.
• 30% say a wrong AI answer has already caused a problem at work; 43% among those who use AI for legal, financial, or compliance questions.
• 51% believe they would personally be legally responsible for acting on a wrong AI answer; 12% believe the AI company would be.
• 52% did not know AI companies’ terms of service disclaim accuracy.
• 22% work under a written employer policy requiring verification of AI output.
• 78% say professionals should be required to disclose AI use to clients.

Methodology

The survey of 500 employed U.S. adults, ages 18 to 64, was conducted September 1, 2026 through the Pollfish online research platform. Respondents were screened to full-time, part-time, and self-employed workers; all 500 reported using AI tools for work at least occasionally. The questionnaire included one occupation question, one usage-frequency question, eight substantive questions, and one attention check, which all respondents passed. Results are unweighted. The margin of error is approximately plus or minus 4.4 percentage points at the 95 percent confidence level for the full sample; subgroup figures carry wider margins.

ABOUT KOLMOGOROV LAW, P.C.

Kolmogorov Law is a California litigation firm based in Irvine, representing businesses and individuals in complex commercial disputes, including breach of contract, fraud, unfair business practices under Business and Professions Code section 17200, computer fraud under Penal Code section 502, trade secret misappropriation, and multi-party commercial litigation across California Superior Courts. The firm is led by Pavel Kolmogorov, who earned his LL.M. from the University of California, Berkeley School of Law, and is licensed in California, the District of Columbia, and multiple federal districts. Learn more at kolmogorovlaw.com.

Media Contact

Pavel Kolmogorov

Kolmogorov Law, P.C.

(909) 235-6420

pavel@kolmogorovlaw.com

View original content:https://www.prnewswire.com/news-releases/half-of-us-workers-know-they-would-be-personally-liable-for-a-wrong-ai-answer-two-thirds-still-dont-always-check-kolmogorov-law-survey-finds-302867898.html

SOURCE Kolmogorov Law