How the we use AI in practice | AI Summit 2026 | Norges Bank Investment Management

| Podcasts | March 24, 2026 | 66 Thousand views | 57:33

TL;DR

Norges Bank Investment Management details its aggressive AI adoption strategy, revealing that achieving a 20% efficiency target required mandatory company-wide upskilling, painful data infrastructure overhauls, and shifting from traditional Scrum teams to small AI-enabled autonomous units.

🏗️ Infrastructure & Data Foundation 2 insights

Insourcing operations and cloud migration

Transitioning from external vendors to owning processes and moving all IT to public cloud removed scalability constraints and eliminated 'data ceilings' that hindered growth.

Consolidating data into Martium Core

Forced migration to a modern data warehouse by setting a hard January 31st deadline ensured high-quality, accessible data essential for AI, despite the tedious work of cleaning legacy datasets.

🚀 Cultural Transformation & Adoption 3 insights

Mandatory upskilling over voluntary programs

Required all employees to complete seven 30-minute AI training sessions because voluntary approaches fail to reach those who need training most, ensuring universal baseline competency.

Ambassador network and Tech Year 2025

Deployed 20 cross-functional ambassadors and embedded AI into every company gathering and summit to create relentless organizational momentum for daily tool usage.

Three-phase implementation strategy

Progressed from distributing tools for bottom-up experimentation to identifying 171 specific efficiency projects, recognizing that widespread small gains outweigh single silver-bullet solutions.

⚖️ Governance & Risk Management 2 insights

Risk-based responsible AI framework

Implemented tiered governance aligned with the EU AI Act, requiring strict human oversight for investment and personnel decisions while applying lighter controls to low-risk systems like email filters.

AI Governance Working Group

Established a cross-functional team to translate compliance rules into actionable processes and adapt governance structures as technology evolves beyond initial training programs.

💰 Business Value & Operational Shift 3 insights

Ambitious efficiency and automation targets

Publicly committed to cutting all manual processes in half by 2028 and achieving 20% overall efficiency gains through systematic AI integration across all departments.

Transforming team structures and rituals

Replaced traditional 8-person Scrum teams with autonomous 3-person units (two developers, one business person) using Claude and Cursor, eliminating ceremonies to accelerate delivery.

AI-assisted block trading decisions

Teams now use AI to evaluate approximately 200 annual block trade requests, analyzing diverse data sources to make rapid decisions on deals like a $30 billion Ferrari stake sale.

Bottom Line

To overcome the 'technology overhang,' organizations must mandate AI training, implement risk-based governance from day one, and aggressively push cultural adoption through small autonomous teams rather than simply providing tools.

More from In Good Company (Nicolai Tangen)

View all
Arvind Krishna: IBM's Reinvention, AI Bets and Quantum | Podcast | In Good Company
57:15
In Good Company (Nicolai Tangen) In Good Company (Nicolai Tangen)

Arvind Krishna: IBM's Reinvention, AI Bets and Quantum | Podcast | In Good Company

IBM CEO Arvind Krishna explains the company's strategic pivot from declining hardware and services to a hybrid cloud and AI software model, detailing the $34 billion Red Hat acquisition rationale, the spin-off of IT infrastructure services, and warning that the current AI infrastructure buildout represents a bubble that only two or three large model providers will survive.

3 days ago · 10 points