By Sofia Beckman – Co-founder, North House: “We manage billions with real-time systems,” one COO told me. “But our governance still lives in PDFs.” That paradox runs through much of the investment management industry. Sophisticated investment strategies run on precise, automated platforms, yet the governance that underpins them often relies on manual routines. Most firms have a policy framework. But governance is not judged by what’s on file. It’s judged by what actually works.
Across the Nordic market, internal governance remains one of the least developed systems inside fund organisations. Not because of missing regulation, but because governance has never been designed as an operational framework. Policies exist, but the systems around them – structure, ownership, and traceability – are weak.
“Most frameworks were written to satisfy regulators at a specific point in time,” Beckman says. “Few were designed to function as live, adaptive structures that can efficiently operate at scale or under stress. That gap has become a strategic risk, and it is still widely underestimated across the industry.”
The weaknesses in governance are no longer theoretical. In 2025, regulators have turned their attention to how control and accountability function inside firms. Supervisors want proof that these mechanisms hold up in practice, not just on paper. Most recent sanctions trace back to shortcomings in governance and internal control. ESMA’s focus on algorithmic management and AI oversight points in the same direction. Governance must now be structured, explainable and verifiable.
With ICT now active under DORA, these expectations have become measurable. Firms are required to show ownership, continuity and incident response in real time. In the Nordics, regulators have already tightened their review of AIF applications, and in Sweden the proposed Lag om olovlig finansiell verksamhet signals a tougher stance on accountability. It introduces criminal liability for unauthorised financial activity and highlights the growing expectation that firms can demonstrate effective internal control. Governance is no longer a compliance exercise. It has become part of the business infrastructure.
Governance breaks down in the details
When North House was founded by Julia Haglind and Sofia Beckman, the two came from opposite ends of the industry. Julia had led Nasdaq Clearing, one of the region’s most system-critical infrastructures. Sofia had worked within a Nordic fund platform, where the operational limits of governance were clear in daily practice.
The challenge was never writing policies. It was connecting them. Understanding how they related, who owned them, and how they were maintained. In most firms, policies lived in silos. Review cycles were manual. Responsibilities were unclear. There was no structural map linking rules/obligations, risks, and accountability.
That missing architecture later became the foundation for the framework developed at North House. Across dozens of hedge funds and AIFs, the same weaknesses repeat:
- Policies stored across SharePoint, Teams, or personal drives
- Review cycles tracked manually via email
- No audit trail of who read or signed off on what
- No link between regulatory requirements and actual controls
“Governance should be the system that links policy to action,” one fund manager said. “Right now, it’s just paperwork.”
Built on structure, not on hype
When ChatGPT launched in late 2022, the North House team was finalising its concept. While much of the industry raced toward AI, the founding team took a different approach. They believed the real problem was not generating more text but managing what already existed.
“The real shift wasn’t the new models,” says Beckman. “It was embedding data structure and policy methodology directly into the design of the platform. Not as an overlay, but as its foundation. That’s what made everything else work.”
Instead of adding new tools, the founding team designed a governance system built on five structural principles:
- Policy as architecture, not archive
Every document follows a hierarchy, from policy to guideline to procedure, making governance readable and enforceable. - Embedded roles and accountability
Ownership, review cycles, and sign-off responsibilities are built in from day one, so policies don’t just exist. They move. - Regulation mapped at the source
Each document is tagged to relevant frameworks, making compliance traceable and auditable. - Data-first, language-ready
Terminology is standardised and metadata machine-readable, creating a foundation for automation and AI. - Continuous oversight
Persona-based dashboards and audit trails give leadership real-time visibility, not just static PDFs.
The results are measurable. In firms adopting this structure, time spent on governance administration has fallen by half. Traceability has reached 100 percent across documents, including edit history and review status. Legal and HR policies can be deployed through ready-made playbooks, cutting onboarding time and ensuring consistency.
The real benefit is not efficiency, but resilience. When structure is in place, automation and AI become useful. Without it, they only add noise.
When structure is in place, AI becomes meaningful
AI alone cannot make governance work. But when ownership, roles, and regulatory mappings are embedded, AI can finally do what it should: support human judgement.
“We don’t treat AI as a surface feature,” says Shamil Limbasiya, AI Engineer at North House. “It’s a logic layer that interacts with structured governance data – version history, ownership, and regulatory tags – so outputs are accurate and explainable.”
At North House, the structured framework enables responsible use of AI. Version comparison is used to track and interpret policy changes, helping teams prepare decision materials and board documentation with full transparency. The same capability supports summaries and translations for audits and investor due diligence, ensuring that every update is traceable and clearly understood across languages and jurisdictions.
Building on this structure, North House is developing a governance assistant that helps teams navigate complex frameworks, summarise key obligations and identify potential gaps that may require review. Nudges notify owners when reviews are overdue, creating a living governance cycle. The aim is to make governance work more transparent and connected while supporting professional oversight.
The foundation that makes everything else work
Governance fails when it is treated as a formality. The firms now moving ahead see it as a system that connects people, processes and information. Hierarchy defines ownership. Clear structures keep accountability alive. Traceability makes decisions defensible.
When governance works this way, it does more than satisfy regulation. It builds stability and speed into daily operations. Decisions become easier to justify because the reasoning is visible. Risks are easier to manage because responsibilities are clear.
Most firms still talk about digitalisation, yet few begin with governance. It remains the missing foundation beneath every process, every role and every decision. The next phase will not depend on new regulation, but on design. Governance built to function, not just to be filed.
“Once governance is structured, it behaves like any other core system,” says Beckman. “It can scale, adapt and integrate. It stops lagging behind strategy and starts enabling it.”
About North House
North House helps fund managers turn governance into a working system. Its Governance Foundation combines structure, ownership, and lifecycle management, transforming fragmented policy frameworks into a connected, operational model.
The platform supports continuous compliance with GDPR and DORA. It is built on Microsoft Azure with data stored in Sweden, and integrates naturally with existing workflows in Word and SharePoint.
