Using personal AI accounts for work tasks is becoming common in healthcare, academia and industry. On 16 November, the Centre for AI Innovation at Karolinska Institutet hosted a fully booked lunch seminar at Innomedicum on how organisations should handle what is known as Shadow AI.

Shadow AI is what happens when researchers, healthcare professionals and other employees use AI tools outside their organisation’s approved systems, often with their own personal accounts. The tools can save time and support everyday work. They can also raise questions about data protection, patient safety and accountability.
Sharing experiences from research and healthcare
The discussion was vivid. Many participants shared their own experiences of Shadow AI in their workplaces, from healthcare organisations and universities to private companies.
Their examples showed that Shadow AI is already part of everyday working life across sectors, and that many organisations are still working out how to meet the need for AI tools while keeping the risks under control.
Measuring traffic to AI tools
- Fredrik Carlsson opened the seminar. He is a medical doctor, senior software engineer and PhD student in AI in healthcare, and part of the Centre for AI Innovation team. His presentation covered what Shadow AI is, why it is spreading and what challenges it creates.
- Markus Lingman, Chief Medical Intelligence Officer at Sahlgrenska University Hospital and a member of the Steering Group, shared his experiences and insights after measuring the traffic to AI tools from employee’s laptops.
- Magnus Boman, Professor in AI within Health at KI and Chair of the Centre’s Steering Group, then led an open discussion.
The discussion looked at notable examples and incidents, the possible benefits of Shadow AI, and how organisations can respond in a constructive way.
More seminars ahead
The Centre for AI Innovation is planning more seminars on AI, innovation, research and healthcare.