Fredrik Standahl in E24 on shadow AI in Norway

E24 published a Fredrik Standahl op-ed on May 19, 2026. The piece argues that shadow AI is now a defining data-exposure pattern in Norwegian workplaces. Employees adopt unsanctioned LLM tools faster than IT departments can sanction them, and sensitive data leaves the company perimeter in the process.

The argument leans on two numbers. In 2025, 56 percent of Norwegians used generative AI. Globally, MIT found in 2025 that only 40 percent of companies have a paid LLM subscription, while 90 percent of employees in the same companies use personal AI accounts for work tasks. That gap is shadow AI.

“The biggest risk is not using AI. It is letting employees use AI alone in the dark.” Fredrik Standahl, in E24.

The piece names two reference incidents. The Lovable platform breach exposed unsanctioned use across Samsung, Amazon and major financial institutions. Samsung engineers previously uploaded confidential code to ChatGPT for debugging, and the data ended up in the model’s public knowledge base.

Fredrik’s recommendation is direct: treat AI as critical infrastructure, with the same control posture as bank transfers and HR systems.

Read the full op-ed at E24.