Israeli Supreme Court Rules Professionals Must Verify AI Outputs, Not Blame Technology for Errors
In a recent Israeli Supreme Court case involving OpenAI's ChatGPT 5, it was revealed that dozens of cited references and rulings were fabricated by the AI. Justice Gila Kanfi-Shtainitz used the opportunity to clarify that even if AI generates false sources, the professional using them bears the responsibility to verify their accuracy. The court ruled that blaming AI for errors cannot serve as a defense against professional liability. Although the case concerned a lawyer, its implications extend broadly across professions.
A May 2026 survey by the Bank of Israel and the Israeli Democracy Institute found that 57% of Israeli workers use AI tools at least weekly, spanning fields such as programming, academia, law, accounting, marketing, engineering, and human resources. AI has rapidly transitioned from a technological novelty to a routine work instrument. However, Israeli labor law has yet to explicitly address who is liable when errors arise from reliance on AI: the employee for failing to verify outputs, or the employer for encouraging AI use without clear guidelines.
Current legal duties of care, loyalty, and professionalism were developed in a human decision-making context, but AI introduces risks of inaccurate results and potential privacy breaches. Mistakes stemming from AI use can lead to disciplinary actions or termination, making the distinction between individual professional error and organizational failure to implement AI responsibly critical.
International organizations have begun adopting internal AI policies that prohibit inputting confidential data into open systems, require human verification of AI outputs, and affirm employee accountability. In contrast, most Israeli workplaces lack uniform policies, leaving employees to adopt AI tools without formal instructions. For example, Microsoft Israel mandates comprehensive security training and provides guidelines on safe AI use, reflecting employer awareness of potential errors and shared responsibility for prevention.
As AI boosts productivity and saves time, employers may face increasing expectations to establish clear usage rules and define tasks suitable for AI assistance versus those requiring human oversight. Future workplace safety and compliance training might also include AI literacy, shifting liability from solely employees to shared employer-employee responsibility. The article was authored by attorney Noa Rubin of Gideon Rubin & Co., specializing in labor law.
