Avoki CEO outlines three rules for workplace AI policies

AI use is spreading across Swedish workplaces faster than clear rules are being adopted, according to a Novus survey commissioned by Avoki. The survey found that 76% of white-collar employees at private small and medium-sized companies use AI tools at work, while 32% said their workplace has clear rules or guidelines. Nearly one in three AI users said they had used a tool without knowing whether their employer had approved it.

Avoki CEO Peter Uddfors says organizations should first find out which tools employees are already using, including services paid for individually, then offer approved alternatives. He argues that policies should be short and straightforward, and sets out three core rules: check AI-generated material before using it in real work or sending it to customers; make clear what information can be entered into each service; and keep responsibility for decisions with employees. Uddfors highlights trade secrets, customer information and personal data as information requiring care. In the survey, 18% of AI users said they had entered work-related information into an AI tool using a personal account.

Training also matters, he says, so employees understand the rules and how to use the tools. The survey found that 64% of AI users apply the tools to less than one tenth of their tasks. Uddfors says larger gains come when organizations redesign processes around AI, rather than focusing first on which tool to buy. He recommends measuring results with existing business indicators, such as handling more customer cases without increasing staffing, shorter lead times, fewer errors, or improved quality and customer satisfaction.

At Avoki, employees were encouraged to experiment, and the company highlighted internal examples to encourage wider use. Uddfors says the most active users are not necessarily managers or technical specialists. The company now uses AI, among other things, to make operational data more accessible and review contracts.