Employee liability for using AI despite a ban

Employers can tell staff that ChatGPT and other generative AI tools are not allowed at the company, but they cannot simply program an automatic disciplinary dismissal for every violation.

Agata Czarkowska, a legal counsel at Hoogells, says employers may set rules for workplace technology when those rules relate to the job and protect legitimate business interests. Polish labour law requires employees to follow work-related instructions that do not conflict with the law or their employment contract. It also requires them to respect workplace organisation, protect the employer's interests and safeguard company information.

Those provisions can support restrictions on AI use. Companies are increasingly concerned about confidential data, trade secrets, information security and the quality of work produced with AI. Some may therefore seek a complete ban or require employees to obtain approval before using such tools.

“

In this company, we do not use ChatGPT or other generative AI tools.

Czarkowska says an absolute ban may not always be the best approach. AI increasingly forms part of software used in everyday work, so employers could instead specify which tools are allowed, what employees may use them for and, most importantly, which data they must not enter. Such rules could appear in an employment contract, although a workplace regulation, security policy or separate AI policy may be a more natural place for them.

If an employee breaks the rules, the consequences require an individual assessment. Termination without notice for employee misconduct requires a serious breach of basic employment duties. The employer must therefore examine which instruction or obligation the employee breached and the circumstances of that case.

“

The employer cannot program a disciplinary dismissal in advance for every click on ChatGPT.