Knut Magnar Aanestad, Legal Process Engineer at Saga, explored how artificial intelligence can create productive friction to improve legal practice quality and accelerate lawyers development. The discussion challenged conventional automation narratives by proposing AI as a tool for intellectual stimulation rather than mere task replacement.

The Technology Optimist’s Journey

Aanestad brought a unique perspective as a Norwegian lawyer who transitioned from traditional practice to legal technology development. His journey began during the previous AI wave in legal services, working with transformer technology since 2017 to develop contract review systems for NDAs and data processing agreements. When ChatGPT emerged, Aanestad leveraged his team’s experience to create “Chat for Legal,” which evolved through mergers with Amsterdam-based companies into what he describes as a “general purpose legal AI platform.”

The platform focused on supporting the lawyering process rather than replacing lawyers. Aanestad emphasised: “It’s not just about the law itself—it’s about lawyering, the highly complex intellectual process. The question is: how can we support that process in the best possible way through a platform?”. This approach recognises structured patterns in legal work whilst preserving the intellectual rigour that defines the profession.

Three Dimensions of AI Integration

Aanestad categorised AI’s influence on legal work into three distinct approaches. Automation represents the most commonly discussed application, where AI completes tasks independently and lawyers review outputs for accuracy. However, this risks reducing practitioners to “quality assurance managers”—a characterisation that highlights the profession’s resistance to purely mechanical roles.

The second dimension involves AI assistance for structuring information, creating timelines, and providing analytical overviews. This collaborative approach maintains lawyer engagement whilst leveraging AI’s computational advantages for data organisation and pattern recognition.

The third and most innovative dimension focuses on creating intellectual friction. Rather than eliminating challenges, AI deliberately introduces productive tension by challenging lawyer reasoning, highlighting weak arguments, and suggesting alternative perspectives. This approach transforms AI from a labour-saving device into an intellectual sparring partner.

Friction as an Educational Tool

The friction concept emerged from Aanestad’s practical experience using his platform’s “Devil’s Advocate” assistant. When challenging a personal legal complaint he had drafted, the AI’s critique forced him to strengthen his argumentation significantly. “What happened was that I needed to sit down and say, yeah, I thought that was a really strong argument. It’s not really, but if I rephrase it a bit like this and that, and then connect it to this report… then it gave more sense and it was a lot stronger.”

This exemplifies friction’s educational value. Unlike traditional continuing education delivered through seminars or conferences, AI-enabled friction operates continuously within daily practice. Every document, argument, or legal analysis becomes an opportunity for intellectual development and skill refinement.

The approach addresses a fundamental challenge in legal education: how to provide experienced practitioners with meaningful intellectual challenge. Traditional methods often fail to create the sustained, contextual engagement necessary for advanced skill development. AI friction fills this gap by offering personalised, immediate feedback within actual work contexts.

Implementation Without Complexity

Contrary to current industry emphasis on sophisticated agents and complex systems, Aanestad advocated for simple implementation approaches. Basic prompts requesting critical analysis of legal work can generate substantial improvement without extensive technical infrastructure. The key lies in designing interactions that challenge rather than merely assist.

This accessibility democratises advanced AI benefits for smaller firms and individual practitioners. Rather than requiring substantial technology investments, lawyers can begin experimenting with friction-based approaches using existing AI tools and straightforward prompting strategies.

Language and Global Accessibility

One attendee highlighted the ongoing challenges for non-English legal systems. However, Aanestad referenced recent research suggesting large language models develop conceptual understanding independent of language before generating responses. This suggests improving multilingual capabilities as models advance.

The conversation revealed varying adoption patterns within firms. Mid-level lawyers (around 30-35 years old) showed greatest enthusiasm for AI integration, whilst junior lawyers remained focused on mastering traditional skills and senior partners approached change cautiously. This generational dynamic influences implementation strategies and training approaches.

Transformation Beyond Technology

Aanestad emphasised that successful AI integration requires comprehensive transformation rather than simple tool adoption. “It’s not only the firm transformation, it’s the transformation of how each and every lawyer works. And that’s where it starts.” This perspective acknowledges the cultural and procedural changes necessary for meaningful AI integration.

User adoption strategies have evolved from brief training sessions to sustained transformation support involving specialists working with firms over extended periods. This reflects the complexity of changing ingrained professional habits and the time required for genuine skill development.

Educational Institution Integration

The discussion touched on extending friction-based AI approaches to legal education. Universities and law schools represent crucial environments for developing AI-enhanced lawyering skills from the beginning of professional training. Early exposure to productive friction could prepare students for practice environments where AI partnership becomes standard.

This educational integration addresses concerns about training future lawyers in an AI-enhanced environment whilst maintaining the intellectual rigour that defines legal excellence.

Strategic Considerations

The friction approach requires careful calibration to avoid overwhelming practitioners or undermining confidence. Effective implementation balances challenge with support, ensuring AI criticism enhances rather than paralyses decision-making.

Firms must also consider client expectations and regulatory environments when implementing friction-based approaches. Whilst stronger legal arguments benefit all stakeholders, the transition period requires managing varied technological sophistication levels across the legal ecosystem.

Conclusion

Aanestad’s friction-based approach represents a sophisticated response to AI’s role in legal practice. Rather than viewing technology as a threat to professional identity or merely a tool for efficiency gains, this perspective positions AI as an intellectual enhancement mechanism that strengthens rather than replaces lawyer capabilities.

The approach acknowledges lawyers’ fundamental desire for intellectual challenge whilst providing practical methods for continuous skill development. By embedding learning opportunities within daily practice, friction-based AI integration offers a path toward enhanced professional capability without sacrificing the analytical thinking that defines legal excellence.

As the legal profession continues navigating AI integration, approaches that respect and enhance human intellectual capacity whilst leveraging technological advantages may prove more sustainable than those emphasising replacement or pure automation. The friction model provides a framework for this balanced integration, suggesting that the future of legal practice lies not in choosing between human or artificial intelligence, but in creating productive partnerships that elevate both.