Legora General Counsel: How AI Turns Legal from a Bottleneck into a Business Partner

From an Impossible Task to Everyday Operations

In an interview, Legora General Counsel Emma Kadri Bergström recalled a late night ten years ago: she was asked to compile the license terms of a batch of supplier contracts overnight. She quickly realized this was an impossible task at the time. Today, with existing tools, this kind of work has become feasible.

This contrast forms the basic framework for how she understands the value of AI—not replacing lawyers, but freeing them from large amounts of low-value document processing so that speed and quality can improve simultaneously. She describes her daily working state as being able to move tasks forward faster, handle more threads at once, and maintain a high quality standard. Through Legora, she can collaborate with colleagues, counterparties, and other third parties on the same platform, with AI participating in the process.

NDA Automation: From 20% to Less Than 1% of Staffing Input

The most concrete example Bergström gave is the handling of non-disclosure agreements (NDAs). She readily admits that NDAs are the kind of document that is mandatory but that no one wants to spend time negotiating, which made them one of the first things the team automated.

Before building the playbook and automation workflow, the team spent about 20% of full-time employee hours on NDA review. After automation, that proportion dropped to less than 1%. The approach was not complicated: first complete the automation, then establish a playbook so that AI performs a first-round review of contracts and flags clauses that fall outside the acceptable range.

The freed-up time was reallocated to work with greater strategic value—larger-value contracts and more strategically significant transactions. Bergström believes this is precisely where the team truly creates the greatest difference, rather than reviewing standardized documents one by one.

What playbook means in contract automation

In the context of contract legal automation, playbook does not refer to a game guide, but to a predefined rulebook: it codifies the team's commercial position and risk appetite for a certain type of contract (such as an NDA), clarifying which clauses are acceptable, which need modification, and which must be escalated to a lawyer for manual handling. A playbook usually includes a clause checklist, acceptable ranges, alternative wording templates, and escalation thresholds. With it, AI can judge whether a contract is "out of bounds," rather than giving a vague good-or-bad assessment. The quality of the playbook directly determines the accuracy of automation—the clearer the rules and the more specific the boundaries, the fewer false positives and false negatives there are in AI's first-round review. Therefore, building a playbook is often more time-consuming than adopting an AI tool itself, and it requires more input from senior lawyers' experience.

How a Small Team Supports Rapid Expansion

Legora's own growth pace provided a stress test for this approach. According to Bergström, over the past 12 months, the company expanded from 3 offices to 18, completed three rounds of financing, and made five acquisitions. Supporting this scale of business with a very small legal team would be almost impossible without the right tools.

She calls Legora the team's "secret weapon" for keeping up with the company's pace and moving projects forward. The key here is not how intelligent the tool itself is, but that it allows the legal team to remain responsive while the business is running at high speed, rather than becoming a bottleneck in the process.

What AI Cannot Replace: Accountability and the Human Side

In Bergström's view, two things in the lawyer's role will always belong to humans. The first is accountability for legal advice—no matter through which channel or with which tool the advice is conveyed, there must be a responsible person behind it. The second is the human side: when lawyers do not have to review documents one by one, the time freed up can be used for strategic planning, meetings with business stakeholders, and aligning goals.

This is also the core logic behind her positioning of legal as a "strategic partner" rather than a "support department." Only when people are freed up is there room to make judgments, coordinate, and maintain relationships—and these are precisely the parts that AI currently cannot take on.

Why accountability in legal advice is hard to transfer to AI

In the legal profession, accountability refers to practitioners being responsible for the professional and legal consequences of the advice they provide. This responsibility is tied to lawyer qualifications, professional regulations, and professional insurance, and cannot be delegated to a tool. Even if AI generates a draft of the advice, signing and endorsement must still be completed by a qualified person. This constraint has two practical consequences: first, AI output must be traceable and explainable so that lawyers can be accountable for it; second, workflow design must retain human review checkpoints and cannot treat AI-flagged results directly as final conclusions. This is also why many legal teams position AI as a "first-round reviewer" rather than a "decision-maker"—what it compresses is the time for searching and comparison, not judgment and responsibility itself.

The Legal Department's Biggest Fear: Becoming a Bottleneck

One noteworthy insight from the interview concerns the discussion of "bottlenecks." Bergström points out that every legal department worries about becoming a business bottleneck. And once it truly becomes one, the consequences are more serious than imagined: business stakeholders will bypass legal and stop coming to you, which instead raises risk and makes lawyers' work harder.

Therefore, "not becoming a bottleneck" is a common driving force for all kinds of legal teams. What she is most proud of is that the team has consistently maintained its position as a business strategic partner—legal is brought in early in projects, enabling it to respond quickly and provide value. This model of early involvement is difficult to achieve without tool support.

For peers who are still watching, Bergström's attitude is open: the technologies emerging now make the legal profession more interesting, and the future is bright. Her advice is direct and simple—try to embrace new technologies and new ways of working, and enjoy the process.

Why "business bypassing legal" raises risk

When legal responds too slowly, business teams, in order not to delay progress, often move contracts or projects forward on their own first, then supplement with legal review afterward, or even skip it entirely. This approach formally achieves business goals, but leaves legal risk in a link where no one reviews it: clauses may be missing, liability allocation may be unbalanced, and compliance requirements may be overlooked. More troublesome is that once this kind of bypass becomes a habit, it is very hard to reverse. Legal is gradually excluded from the decision-making chain, and by the time problems surface, the window for remedy is often already missed. Therefore, "not becoming a bottleneck" is not only an efficiency demand but also a risk management demand—the visibility and early involvement of legal are themselves part of an organization's risk control capability.

One-line Summary

Legora's General Counsel discusses how AI reduced NDA review hours from 20% to less than 1%, making legal a strategic partner rather than a bottleneck.

Paragraph Summary

Legora General Counsel Emma Kadri Bergström uses her own experience to explain that the value of AI lies not in replacing lawyers, but in freeing them from low-value document processing. The most concrete case is NDA review: through automation plus a playbook, AI performs a first-round review and flags out-of-bounds clauses, reducing the team's full-time employee hours from about 20% to less than 1%, and the freed-up time is reinvested in larger-value, more strategically valuable transactions. She emphasizes that lawyers' accountability and interpersonal collaboration will always belong to humans, and AI cannot take them on. The biggest risk for legal departments is becoming a bottleneck that causes business stakeholders to bypass them and raises overall risk, so early involvement in projects and maintaining a strategic partner position are key.