Holtman Systems Research

Holtman Systems Research is a Dutch SME founded by Koen Holtman in November 2022 and based in Eindhoven, the Netherlands. It operates as a single-person company with the mission to benefit society by doing research and applied innovation on improving the safety of AI systems and improving interactions between AI systems and society. Koen Holtman is a computing scientist and systems architect with a PhD in Software Design from Eindhoven University of Technology, earned through computer science research conducted at CERN in Switzerland. He brings over 20 years of experience in industrial research and development, including work at Philips, and over a decade of experience in standards creation. His earlier standards contributions include co-authoring RFCs defining the HTTP/1.1 protocol and contributing to the Blu-ray Disk system specifications. Holtman began independent AI safety research in 2019, publishing papers on arXiv covering agent foundations, corrigibility, and AI alignment. Key publications include Corrigibility with Utility Preservation (2019), Towards AGI Agent Safety by Iteratively Improving the Utility Function (2020), Counterfactual Planning in AGI Systems (2021), and co-authoring Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systems (2024). In 2023, Holtman transitioned to full-time AI governance work. The company's primary current activity is participating in the CEN-CENELEC JTC 21 committee, which is writing European AI safety standards in support of the EU AI Act. Within JTC 21, Holtman serves as a member of the WG2 Risk project Editorial Team and co-convener of the WG1 task group on horizontal/vertical standardization. He is most active in the WG2 project writing the European AI Risk Management standard. Holtman also co-founded the AI Standards Lab, where he serves as co-lead and resident standards writing expert. The AI Standards Lab is a Delaware nonprofit (fiscally sponsored by Players Philanthropy Fund) with a mission to accelerate the writing of AI safety standards by bringing together and supporting independent experts. The Lab has a small team of research analysts and operations staff working on AI governance research and standards development. Holtman Systems Research has received funding from the Survival and Flourishing Fund across multiple grant rounds, supporting both its independent standards work and its role in the AI Standards Lab. The company also offers consulting services in AI technology, AI regulation, and socio-technical systems safety, though it is not currently seeking new customers.