From metal behemoths to agentic AI - what Scotland's legal sector can learn from bridging innovation

Dr Corsino San Miguel PhD, LLB - a member of the AI Research Group and the Public Sector AI Task Force at the Scottish Government Legal Directorate - considers Scotland's innovation past and what the profession can learn for the future.

Progress in any field is seldom linear; it is shaped by necessity, driven by innovation, and guided by vision—but never reliant on perfection. Few countries embody this truth as distinctly as Scotland. The history of its bridges exemplifies why embracing technology early, even in its imperfect forms, is essential to catalysing growth and enabling future advancements.

The Forth Bridge, built with a pioneering combination of steel construction and the cantilever principle, redefined engineering standards and emerged as a global symbol of innovation. By the 1960s, demand for greater connectivity and the irruption of another transformative technological advance—the car—prompted the construction of the Forth Road Bridge. This, in turn, led to the Queensferry Crossing in 2017, built to overcome the Forth Road Bridge's shortcomings. These later bridges weren’t built solely because technology had advanced—they were built because necessity demanded it. The Forth Bridge, once a symbol of progress, could no longer sustain the increased traffic and economic activity it had helped create. Without the initial step of building the first bridge in 1890, the subsequent development and demand for modern infrastructure might never have materialized.

This sequence of progress offers a valuable lesson for the Scottish legal profession's journey with AI. Some argue that AI is too immature for widespread adoption in law, but history shows us that waiting for perfection is a misstep that hinders progress. If early bridges hadn’t been built, there would have been no growth to necessitate and justify better bridges later. Similarly, law firms that fail to adopt AI now—despite its imperfections—risk stagnation. Whether you’re reading this article today, in 2026, or in 2030, AI technology is at its least advanced point right now—it is only growing stronger.

2025 marks a pivotal moment in the evolution of AI, as the industry transitions from generative to agentic intelligence. This shift signals an era where AI advances beyond simply generating insights to actively executing complex, multi-step tasks with end-to-end autonomy. Leveraging sophisticated reasoning models capable of verifying their outputs, AI is poised to move from standalone chatbots to seamlessly integrated systems that replicate and elevate the workflows of legal professionals. Rather than "buying assistants," we are entering an ecosystem where we are effectively "hiring coworkers".

Agentic AI combines the adaptability of large language models (LLMs) with the precision of traditional programming, creating systems capable of autonomously performing tasks, designing workflows, and leveraging tools to make decisions and solve complex problems. These systems go beyond their training data, drawing from databases, networks, and user behaviour to continuously improve over time. By orchestrating the behaviour of multiple agents—ranging from additional AI models to real-time search tools—agentic AI platforms enable vertical dynamic and personalized interactions across diverse applications.

Driving this transformation is the emergence of groundbreaking reasoning models like OpenAI’s o1, launched in September. With its "chain of thought" methodology, the o1 model excels at breaking down intricate problems into actionable steps, enabling accurate and autonomous execution of complex, multi-step tasks. Its seamless integration into diverse workflows not only enhances operational efficiency but also frees users to focus on higher-level decision-making and strategy. By redefining AI’s role from a passive tool to a collaborative partner, the o1 model exemplifies the promise of agentic AI in advancing the legal profession.

The transition from generative AI to agentic AI marks a pivotal shift in how AI interacts with the legal profession. While generative AI has elevated the importance of structured legal databases, agentic AI redefines the conversation by prioritizing workflows and tacit knowledge over raw data. Just as an excellent chef’s artistry lies not merely in the recipe or ingredients but in their unique processes and know-how that make each dish exceptional, so too does the legal profession’s strength lie in its workflows. This evolution positions law firms—not database owners—as the new leaders in the AI landscape, leveraging their unique expertise to build systems that mirror their distinct practices.

The power of agentic AI lies in its ability to execute complex, multi-step processes autonomously. Unlike generative AI, which primarily focuses on generating insights from data, agentic AI thrives on structured workflows, reflecting the nuanced decision-making and judgment inherent in legal practice. For instance, drafting a contract is not merely about accessing precedents; it involves understanding context, applying strategic thinking, and iterating on feedback. These tacit elements—often undocumented but integral—define the legal profession and give agentic AI its edge.

Workflows encapsulate more than a sequence of tasks; they embody the expertise, strategy, and problem-solving skills developed over years of practice. By embedding these workflows into agentic AI systems, law firms create tools that not only replicate but also enhance their methodologies. Ultimately, the rise of agentic AI signals a new era where the expertise embedded in workflows becomes the defining factor of AI’s success in law. By shifting the focus from data to execution, agentic AI not only reshapes the role of technology but reaffirms the centrality of law firms as innovators and leaders in the AI revolution.

As outlined over a year ago in Navigating the AI Frontier, I argued that the middle market would become the critical battleground for the future of legal services in Scotland. While small firms benefit from agility and cost-effective AI solutions, and large international firms leverage resources to deploy advanced AI technologies, mid-sized independent firms face unique challenges. There is no time for luxury scepticism or prolonged hesitation—success will depend on decisive action. Agentic AI is not just a technological upgrade; it is a paradigm shift that demands new thinking, focusing on embedding workflows and tacit knowledge to drive innovation and efficiency.

To effectively confront the emerging challenges of the legal profession in the age of agentic AI, middle size independent Scottish firms must prioritize leveraging their proprietary knowledge and structuring the tacit expertise embedded in their practices. Proprietary data—spanning decades of legal documents, case law analysis, and templates—is a unique asset that firms can use to fine-tune AI models tailored to their specific needs and values. However, the real transformative potential lies in capturing and formalizing tacit knowledge: the nuanced workflows, strategic thinking, and decision-making processes that define a firm's excellence. By converting these intangible elements into structured, AI-trainable formats, firms can create systems that not only replicate their methodologies but also enhance them. This dual focus on proprietary and tacit knowledge positions firms to navigate complexity with precision, maintaining a competitive edge while seamlessly integrating innovation into their legal practice.

The risk posed by agentic AI is not just external but deeply internal. If law firms aren’t the first movers, individual lawyers might take the lead. Legal tech firms are aggressively hiring seasoned lawyers to develop cutting-edge capabilities, leveraging their expertise to reshape the competitive landscape. Yet, the more profound risk lies within the firms themselves: if partnerships fail to recognize the transformative potential of capturing and formalizing tacit knowledge—those nuanced workflows, strategic insights, and decision-making processes that define a firm's excellence—they risk losing their competitive edge. By neglecting to convert these intangible elements into structured, AI-trainable formats, firms leave the door open for individual partners, in collaboration with AI developers, to achieve this independently. Such moves could fragment the firm's expertise and transfer its intellectual capital outside its traditional boundaries. To remain relevant, firms must embrace this challenge, harnessing their collective knowledge to build AI systems that enhance their methodologies and secure their position in an increasingly technology-driven legal landscape.

The history of Scotland’s iconic bridges teaches us that progress is born not from waiting for perfection, but from daring to innovate with what is possible today. The legal profession now stands at a similar juncture with agentic AI, a transformative leap that redefines the relationship between technology and practice. By embedding workflows, tacit expertise, and strategic thinking into AI systems, Scottish law firms have the opportunity to lead this new frontier. The path forward is clear: those who act decisively will not only weather the tides of change but shape the future of the profession itself. Much like the bridges that define Scotland's progress, the AI systems we build today will be the foundations upon which tomorrow’s legal innovation thrives.

Article written by Dr Corsino San Miguel PhD, LLB in Scots law and graduate in Spanish law, co-founded and led European Telecom Company before entering academia. He is now a member of the AI Research Group and the Public Sector AI Task Force at the Scottish Government Legal Directorate.

The views expressed here are personal.