When AI is integrated into operational functions, it places demands on how the companies concerned understand and manage risk. Havtil is currently conducting a series of audits focused on risk management and the technology qualification of new technologies such as AI among various operators.
In a recent audit of Aker BP, Havtil examined the AI technology that is to be introduced at Yggdrasil. The objective of the audit was to verify that Aker BP had established and applied requirements and systems for risk management and qualification to ensure safe operation.
Yggdrasil is particularly relevant because the project involves a high degree of automation and digitalisation. Control and monitoring will largely be carried out from shore, while the operation will have a reduced or no permanent crew offshore.
At Yggdrasil, Aker BP plans to use robots that can be remotely controlled from shore. The goal is for certain robots to gradually take over operator inspection rounds and, for example, detect gas leaks, heat generation and other abnormal conditions. There are also plans for fully electric cranes that can be remotely controlled from onshore. In this case, AI models will provide information to the crane operator during normal operations and help detect and manage incidents.
Read the audit report in full here:
Aker BP – Yggdrasil – risk management and qualification of new technologies such as artificial intelligence
Tracking developments over time
Over time, Havtil has been studying how AI can impact safety in the petroleum industry.
Among other things, in 2025, we sent a guidance letter to the industry explaining how the regulations may apply with regard to AI. As part of our audit of Aker BP, we followed up a range of issues, including how the contents of the letter were communicated and evaluated.
In 2026, Havtil also mapped how the operating companies are following up on AI. The survey revealed varying levels of maturity and a need for a more systematic approach to areas such as oversight and risk management.