If you have been following developments in regulatory technology, you may have noticed something. The term norm engineering is starting to appear more often in conversations about how regulation can function in a digital world.
It is not yet mainstream, but it is gaining momentum. At the same time, the idea itself is not new.
For decades, researchers have explored how legal reasoning could be translated into forms that computers can process.
Early work on legal expert systems in the 1980s focused on structuring rules and decisions, laying the foundation for machine-readable law. Over time, advances in knowledge engineering made it increasingly possible to model complex domains such as regulation in a consistent way.
By the 2000s, the ambition was clear. If regulation could be expressed in a structured, machine-readable form, it could improve policy design, reduce compliance costs, and make decisions more transparent and consistent.
The promise was visible, but it did not scale.
For years, norm engineering remained limited to specialized projects and expert teams. It delivered value, but it did not spread across organizations or industries.
So, what held it back?
What held norm engineering back
To understand what held norm engineering back, it is important to first understand what it set out to achieve.
At its core, the idea is straightforward. Regulation, as it is written today, is designed for human interpretation. Laws are expressed in natural language, which makes them flexible, but also complex and difficult to apply consistently at scale. This creates challenges for both those who must comply with regulation and the institutions responsible for interpreting and enforcing it.
Norm engineering takes a different approach. Instead of treating regulation as text to be interpreted each time, it focuses on translating legal rules into structured, machine-readable representations that can be applied consistently by systems. This makes it possible to capture conditions, obligations, and decisions in a way that can be processed by systems.
When regulation can be expressed in this structured form, it becomes possible to automate parts of decision-making, support policy design with simulation, and provide clear explanations for outcomes. This has the potential to improve the quality of regulation while reducing the cost and complexity of compliance.
This vision has been consistent over time. From early expert systems to more recent developments in regulatory technology, the goal has remained the same: to make regulation more accessible, transparent, and reliable in its application.
In that sense, norm engineering did not fail because the idea was flawed. It struggled because the conditions required to realize that idea at scale were not yet in place.
Three bottlenecks stood in the way:
- The skill gap
- The absence of shared standards
- Low productivity
Bottleneck 1: The skill gap
The first bottleneck was a matter of people.
Early norm engineering efforts depended on a rare combination of expertise. It required “unicorn” individuals who understood legal texts, could interpret their meaning in a specific domain, and were also able to translate that interpretation into structured models that systems could process.
In practice, these capabilities did not often exist in a single role. Legal interpretation, domain expertise, and technical modeling have traditionally existed as separate disciplines, each with its own methods, language, and training. Bringing these perspectives together required coordination across multiple roles, which added complexity and made norm engineering difficult to scale.
Projects depended on scarce expertise, and progress was often limited by the availability of people who could operate across these domains. As a result, even when the value was clear, expanding beyond specific use cases or organizations remained a challenge.
This challenge is reflected more broadly across industries. Among firms reporting skill gaps, 63% face increased workloads for existing staff, 50% report higher operating costs, and 46% struggle to implement new working practices. One in three say skill gaps limit their ability to adopt new technologies.
Bottleneck 2: The absence of shared standards
The second bottleneck was a lack of shared standards.
While organizations explored ways to model and interpret regulation, there was no common approach to representing legal meaning in a structured and reusable form. Each implementation developed its own methods, definitions, and interpretations, which limited reuse and made collaboration difficult.
Without shared standards, models could not easily be exchanged or scaled across teams, organizations, or jurisdictions. Efforts remained isolated, and the same work often had to be repeated from one implementation to another.
Bottleneck 3: Low productivity
The third bottleneck was productivity.
Translating regulation into structured, machine-readable models required significant time and effort. Interpreting legal text, aligning it with domain context, and formalizing it into a usable structure was a largely manual process, making it difficult to scale across large regulatory domains.
This slowed adoption. Even when the long-term value was clear, the upfront investment required to model regulation created a barrier. As regulatory complexity continued to grow, so did the effort needed to keep models accurate and up to date.
Why these bottlenecks are finally breaking
For years, these bottlenecks defined the limits of norm engineering, preventing an idea with clear value from scaling beyond specialized use cases.
That is now starting to change as organizations rethink how they approach regulation and compliance. What was once treated as a reactive, isolated function is becoming more integrated, more strategic, and more central to how organizations operate. In a 2023 Thomson Reuters Risk & Compliance survey, 70% of respondents report moving away from check-the-box compliance toward a more strategic approach over the past two to three years.
This shift is not happening in isolation. It is being driven by a combination of technological and organizational changes. It is also reflected in the growing ecosystem of practitioners and initiatives such as the Norm Engineering Conference 2026, which bring together experts working to advance the field.
1. Collaboration platforms are reducing the skills gap
Organizations are moving away from relying on rare, hybrid expertise and toward collaborative ways of working. Legal experts, domain specialists, and engineers can now contribute within shared environments, each using their own language and perspective. This reduces dependency on “unicorn” profiles and makes it easier to scale knowledge across teams.
2. Open standards are enabling reuse and interoperability
Efforts to establish common structures for representing regulation are gaining traction. As interoperability becomes a priority across legal and regulatory systems, organizations are beginning to align on shared approaches, making it easier to reuse models, exchange interpretations, and build on existing work.
3. AI is transforming productivity
Advances in artificial intelligence are dramatically reducing the time required to interpret and structure regulatory content. Tasks that once required extensive manual effort can now be supported or accelerated, allowing organizations to move faster and operate at scale. This reflects a broader trend, as companies increasingly adopt AI to improve efficiency and decision-making across operations.
As explored in our earlier article on the future of AI-enabled norm engineering, the combination of structured regulatory knowledge and AI is enabling this transition from manual interpretation to systems that support execution at scale.
Together, these developments are removing the barriers that once limited norm engineering.
From possibility to scale
For years, norm engineering remained an idea with clear promise but limited reach, held back by limitations that made it difficult to scale. That is now changing as collaboration improves how expertise comes together, standards begin to align how regulation is represented, and advances in AI accelerate how quickly it can be applied.
This shift goes beyond efficiency and starts to reshape how organizations interact with regulation. Instead of repeatedly interpreting rules, they can begin to apply them consistently through structured, scalable approaches. Instead of reacting to change, they can design for it, embedding regulation directly into systems and processes.
Norm engineering has always promised value at scale, but for the first time, the conditions are finally in place to realize that promise in practice.
