The Legacy Industry Bottleneck
The legal technology revolution has been engineered by and designed for the tech industries, like software companies, financial services platforms, and data-native enterprises. But what happens when cutting-edge artificial intelligence collides with asset-heavy sectors governed by layers of complex regulations, entrenched bureaucracy, and high-stakes physical and judicial liabilities?
The answer, for most legal departments in these sectors, has been a painful paradox: extraordinary pressure to adopt AI, coupled with structural conditions that resist it. Legacy industries carry an enormous volume of legal obligations, municipal permits, environmental licenses, zoning disputes, supply chain contracts, social housing compliance frameworks and a production site that is mostly physical.
In that scenario, basic automation tools are facing difficulties, but the pressure to pivot is coming.
The thesis here is both urgent and actionable: to survive the next decade, General Counsels in traditional, asset-heavy industries must move decisively beyond basic automation and embrace what I am calling “Agentic Legal Ops”.
This shift will require not just the creation of agents and orchestration, but the adaptation of the logic in our model of GRC (Governance, Risk, and Compliance). Yet the technology alone is insufficient. Sustainable success demands a parallel commitment to a human-centric leadership model that embraces cultural transformation.
The Rise of Agentic Legal Ops
There is a critical, misunderstood distinction at the heart of this conversation: the difference between automation and autonomy.
Automation, in the legal context, refers to rules-based tools that execute predefined tasks when triggered by specific conditions. A system that auto-populates a contract template when a deal stage advances in a CRM; a workflow that routes an NDA to the correct approver based on deal value; a dashboard that flags regulatory deadlines on a calendar. These tools have genuine value. But they are reactive, brittle, and fundamentally dependent on humans to define every fork in the road.
Agentic AI, by contrast, refers to systems capable of executing complex, multi-step workflows with a degree of autonomy. Advanced AI extensions are transforming modern Legal Operations by moving far beyond a simple “question-and-answer” function to execute complex, contextual tasks directly within a professional’s browser workflow.
Instead of just drafting standalone text, AI for litigation lawyers can now actively interface with legal platforms to streamline end-to-end tasks: it can automatically navigate and log into judicial systems like the PJe, map out open deadlines from incoming subpoenas, and draft highly localized initial petitions or responses using the firm’s pre-existing templates. Furthermore, it significantly accelerates case analysis and knowledge management by digesting entire lawsuits in minutes, instantly mapping timelines, evidence, and parties while simultaneously scanning databases or the STJ for favourable jurisprudence, summarizing lengthy depositions, and standardizing file management by auto-organizing downloads into specific client folders.
For contract lawyers, an agentic legal system does not simply populate a template; it reads the incoming contract, cross-references its terms against a jurisdiction-specific regulatory database, identifies deviating clauses, scores aggregate risk exposure, proposes redlines with supporting rationale, and flags unresolved issues for attorney review.
For traditional industries, this distinction is transformational as Real estate and infrastructure transactions are document-intensive by nature. A single development project may generate hundreds of contracts across suppliers, public agencies, financiers, and regulatory bodies, each subject to different governing law and compliance obligations. Agentic workflows can move the legal department from passive document storage to active risk management: continuously monitoring contract portfolios, scoring exposure against shifting municipal regulations, and triggering alerts when judicial trends in a specific jurisdiction create material risk in existing agreements.
But the General Counsel who wants to implement this shift must know that the decisive shift from AI assistants to AI agents was not a matter of intelligence; it was a matter of hands. For years, large language models could reason, draft, and analyse, but remained confined to the boundaries of a conversation window, unable to act on the world beyond it. Tools changed that. They gave agents the capacity to do, not merely to advise. This is the architectural inflexion point that separates the assistant era from the agentic one.
That autonomy, however, only becomes governable when agents are connected securely to corporate systems, data repositories, and specialized legal tools through APIs, enterprise connectors, or emerging interoperability standards such as the Model Context Protocol (MCP). It is this secure, structured connectivity that transforms an AI model from a sophisticated drafting aid into a genuine operational actor within the legal department.
The Guardrails of Trust: Governance, Risk and Compliance
The introduction of autonomous AI agents into a corporate legal framework does not eliminate risk, it changes it. General Counsels who move aggressively into agentic architectures without parallel investment in AI governance are trading known legal risks for novel, and potentially more complex, ones.
The rapid integration of AI into the corporate ecosystem is fundamentally transforming the practice of GRC from a framework of static, periodic reviews into a model of dynamic, real-time oversight. A profound competence gap currently exists within boards of directors and executive committees, leaving leadership teams to navigate and govern without the tools to properly decode the risks. But this gap is not only technical, it is also linguistic. Board members and executive committees are fluent in the vocabularies of reputational, financial, and operational risk; AI risk, as it is typically presented, speaks a different dialect entirely. The strategic imperative for the General Counsel, therefore, is not simply to raise AI risk on the board agenda, but to translate it: to reframe model hallucination as reputational exposure, data governance failure as regulatory and financial liability, and over-reliance on autonomous agents as operational concentration risk. When AI risk is mapped onto the risk categories that boards already own and govern, it stops being only a technology conversation.
To mitigate this extreme institutional vulnerability, GRC practices and Audit Committees must urgently step up.
Modernizing corporate governance demands an immediate structural rethink of both board agendas and executive literacy. The Board of Directors must expand its standard oversight to include a new, multi-dimensional matrix that scrutinises five critical pillars: technical and legal defensibility, genuine productivity gains against hidden implementation costs, digital literacy, market distribution impacts, and the continuous monitoring of post-deployment “AI loops.”
For the General Counsel, learning how to help mitigate these complex vulnerabilities and preparing for upcoming AI rules and legislation will be crucial. Policy updates alone will fall short. This journey will require an immediate investment in formal training -such as dedicated AI law courses for the Board, Executive Committee and the GRC team.
A well-designed architecture transforms the attorney’s role from information processor to strategic decision-maker. This is not a concession to caution—it is the correct design for maximising the value of both human and artificial intelligence in a high-stakes legal environment.
Beyond the Human-in-the-Loop, the architectural principle that keeps humans in the review chain of individual AI outputs witch is crucial for Legal frameworks, we have to approach our design with a more ambitious concept: Human-in-Command. The distinction matters. Human-in-the-Loop is an operational safeguard; Human-in-Command is a governance philosophy.
In practice, Human-in-Command means that the General Counsel, supported by cross-functional AI governance committees, acts not as a supervisor of individual AI outputs but as the architect of the framework within which AI is permitted to act. These committees, which should bring together legal, technology, risk, and human capital leadership, are responsible for defining the boundaries: what data agents can access, what decisions they may influence, what thresholds trigger mandatory human review, and what categories of judgment remain permanently outside the scope of automation.
Balancing High Performance with Human Value
The challenge is to drive innovation and technological velocity and, at the same time, promote a human-centric leadership. This is where most digital transformation initiatives in legal departments fail.
Implementing agentic systems into a legacy legal department triggers a predictable and human response: anxiety. Teams that have built careers on mastery of document review, regulatory interpretation, and procedural expertise face a genuine and legitimate question about what their contribution looks like in this new AI environment. When this anxiety is not addressed with directness and empathy, it becomes friction, resistance to adoption, and a fragmentation of the organizational culture that the transformation was meant to strengthen.
The answer to that anxiety is not reassurance; it is architecture. The real design challenge for the organisation is not choosing between technology and people, but building a framework in which both operate at their highest potential: where machines handle the volume, velocity, and pattern recognition that overwhelm human cognition, and where people contribute with ethical reasoning, emotional intelligence, creative thinking, and the irreducibly human capacity to navigate ambiguity with conscience. The goal, therefore, is not an AI-augmented legal department. It is a human-centred one, made more powerful by AI.
The General Counsel’s role in this environment is no longer solely about technical oversight. It is about designing the work and managing a psychological transition. The most effective leaders in this moment are those who can hold both the ambition of the transformation and the humanity of the people living through it.
This imperative is not only a matter of organisational best practice, but it is also increasingly a regulatory one. As an example, Brazil’s evolving labor framework, including the recent updates to NR-1 (the General Occupational Health and Safety Standard), now requires employers to formally identify and manage psychosocial risks in the workplace, including those arising from significant organisational changes. The NR-1 update signals a broader global trend in which psychological safety is migrating from soft HR aspiration to hard regulatory requirement, and forward-thinking General Counsels are already building this into their change management frameworks.
In practice, this means that we need to invest in retraining not as a checkbox but as a genuine repositioning of roles, helping team members understand that the hours reclaimed from document processing are hours available for institutional relationship management, complex negotiation, strategic advising, and the empathetic advocacy that no algorithm can replicate. It means creating visible opportunities for team members to shape the AI systems their department deploys, making them architects of the change rather than subjects of it. And it means modelling, from the top, a posture of curiosity rather than certainty: acknowledging that this technology is evolving rapidly, that mistakes will be made, and that the organisation is committed to learning through them rather than hiding from them.
The legal departments that navigate this duality successfully will be those that emerge from transformation with not only greater operational efficiency but greater organisational cohesion. I believe that teams that trust their leadership because they were treated as partners in the process are going to be more successful in this transition.
The Future-Proof General Counsel
The legal departments that thrive in traditional industries over the next decade will be distinguished by the quality of the frameworks, in both fronts: technical and human.
The future-proof General Counsel in asset-heavy, highly regulated sectors is, necessarily, a hybrid figure: part technologist, fluent enough in agentic architectures to govern them with rigor; part risk strategist, capable of mapping the second-order legal implications of AI deployment before they become liabilities; and above all, a human-centric leader, who understands that the most complex system in any legal department is not the AI platform, but the team of people who must trust it, challenge it, and ultimately take responsibility for the decisions it informs.
This is, in my opinion, the defining condition of legal leadership in traditional industries right now, and the leaders who embrace both sides of it will define what the next generation of the profession looks like.
About the Author
Rafaella Carvalho Corti, General Counsel at Cyrela Brazil Realty
Rafaella Carvalho Corti is a Statutory Director at Cyrela, a leading Brazilian real estate developer, where she heads the Legal, Compliance, Privacy, and ESG departments. With over a decade of experience in corporate governance, she bridges the gap between robust legal strategy, regulatory compliance, and business growth. Rafaella holds degrees in both Law and International Relations, bringing a deeply strategic and international perspective to corporate leadership and risk management.
Beyond her executive corporate role, she is a Professor at Futurelaw, specializing in Strategic Planning and Law 4.0, where she teaches legal professionals how to navigate digital transformation and artificial intelligence. She also serves as the Co-editor of Revista da Carreira and actively contributes to the institutional market as a member of the Legal Committee of ABRAINC (Brazilian Association of Real Estate Developers) and specialized commissions at the OAB/RJ (Brazilian Bar Association). She is an advocate for a leadership model that balances high performance with human-centric management.
