Practical use of AI/LLM

97% Adopted AI. What Now?

Hallvard Nygård

Hallvard Nygård

Architect at Storebrand by day. Making a better society by night. Transparency, openness, and evaluating security and privacy in various systems. Managing Norway's largest database of public records, covering 686 authorities. Holding power to account.

Some previous presentations:

  • BSides Oslo 2024 - Keynote - Exploring the World of Ethical Hacking: From Web Vulnerabilities to Election Security (partial recording - https://www.youtube.com/watch?v=VwWPzRceCgs)
  • HelloStavanger 2024 - 5 sikkerhetsfeil hos Storebrand (not recorded)
  • Javazone 2020 - Overvåke gata eller finne korrupsjon Norges data er ferdig indeksert, https://vimeo.com/showcase/7568031/video/459871711
  • ISF Høstkonferansen 2018 + Meetups 2018 (OWASP Oslo / Javabin Sørlandet) - Walk-through of responsible disclosures I've done
  • Javazone 2017 - Lightning talk - Æ fant data, https://2017.javazone.no/program/a181ba49a1e1450c9a37d0d793a3e4ec
1 / 2

97.3% of developers at Storebrand use AI tools. 81% daily. That was February 2026. Getting here wasn't the hard part.

The hard part is everything after.

We're seeing the way developers work transform — how we collaborate, what we spend time on, what skills matter. The bottleneck has moved from writing code to reviewing it. Team structures are shifting. Nobody wants to go back.

But the questions are piling up: what upskilling do people need? How do team roles and structures change? How do juniors learn the craft? What does the engineering ladder look like when the work itself is transforming?

A field report from inside a transformation that's still happening.