Norway's Sovereign Wealth Fund: Half of Staff Uses Claude for Programming, Plans to Enable AI to Autonomously Participate in Investment Decision-Making

Gate News reports that on March 25th, nearly half of the employees at Norway’s sovereign wealth fund (Norges Bank Investment Management), which manages $2.1 trillion in assets, have been using Anthropic’s Claude large language model to develop AI tools on their own. Stian Kirkeberg, head of machine learning and AI at the fund, introduced at an AI seminar that employees mainly use these tools to assist decision-making, including monitoring ESG and financial risks of 7,000 portfolio companies, simulating contract negotiations, and preparing for company meetings. Kirkeberg stated that in the future, some AI agents will be allowed to make limited autonomous decisions under human supervision. He emphasized that the principle is to let AI analyze data to support better human decisions, and at a certain stage, trust in the AI to make some decisions, but human oversight remains essential. This model has not yet been implemented. The fund’s CEO, Nicolai Tangen, previously said that companies not adopting AI are “complete morons.” He revealed that the fund has invested several million kroner in AI, with returns reaching hundreds of millions of kroner. Currently, the fund uses AI to analyze trading opportunities to reduce transaction costs. The staff size will remain around 700, but roles will shift from back-office administration to front-end investment.