TwinLadder Weekly

Issue #17 | October 2025

---

Editor's Note

A client called last week asking whether they should "get one of those AI contract platforms." When I asked what problem they were trying to solve, there was a long pause. "Everyone says we need one."

That exchange captures where the contract intelligence market sits right now. The technology has genuinely matured — AI algorithms now review commercial agreements in 26 seconds, rendering the manual 92-minute benchmark obsolete. The market is projected to reach $7.2 billion by 2033. But most firms buying these tools still cannot articulate the question the tool is supposed to answer.

That is the real problem. Not which platform. Why. And until you can answer the why with specificity — not "we need to be more efficient" but "we need to identify auto-renewal exposure across our 4,000 supplier contracts before Q2" — the technology investment will underperform.

I see this pattern across Europe with particular frequency. Firms in Berlin, Amsterdam, and Stockholm are purchasing contract platforms because London and New York firms have them. That is not a strategy. That is mimicry. And mimicry at six figures per year is expensive.

---

[HIGH CONFIDENCE]

Contract Intelligence Has Moved Beyond Review — But Most Buyers Have Not

Three years ago, "AI contract review" meant extracting key terms faster than associates could. That capability is now baseline. The leading platforms have moved through three tiers: review (term extraction), repository (centralised storage and search), and intelligence (portfolio analytics, relationship mapping, risk detection). Most organisations have reached the first tier. Few have operationalised the third.

Platform	Niche	Key Metric	Annual Cost
Kira by Litera	M&A due diligence	64% of Am Law 100, 1,400+ clause types, 90%+ accuracy	$50K-$150K+
Ivo	Enterprise in-house	97% accuracy (CUAD benchmark), 6x ARR growth	Enterprise pricing
Evisort (Workday)	Enterprise workflow	286% ROI per Forrester	Bundled with Workday
Market overall	All segments	$7.2B projected by 2033, 23% CAGR	Varies widely

Workday's acquisition of Evisort in September 2024 for an estimated $250-300M signals the direction. Enterprise software companies are absorbing standalone legal AI to add capabilities to their existing suites. Meanwhile, Ivo raised $55M at a $355M valuation, purpose-built for in-house teams, with clients including Uber, Reddit, and IBM. And Kira by Litera remains the M&A due diligence standard, used by 64% of Am Law 100 firms, extracting over 1,400 clause types with 90%+ accuracy.

The honest picture: each platform has carved a distinct niche. Kira dominates M&A review at $50K-150K+ annually. Ivo serves enterprise in-house teams with 97% accuracy on the Contract Understanding Atticus Dataset and 6x ARR growth in a year. Evisort delivered 286% ROI per Forrester — but post-acquisition, its roadmap is now driven by Workday's priorities, not standalone legal users.

What concerns me is the mid-market gap. These platforms are engineered for organisations that can afford six-figure annual commitments and have legal operations teams to manage deployment. A 30-lawyer firm handling commercial contracts has genuine need for portfolio analytics but no path to get there at current pricing. The consolidation wave — Workday buying Evisort, Litera buying Kira — benefits platform ecosystems. Whether it benefits practitioners outside those ecosystems remains unclear.

I have seen this play out in practice. A Fortune 500 manufacturer deployed a contract intelligence platform across 15,000 contracts and discovered 847 auto-renewal clauses triggering within 90 days — many for services no longer used. Estimated annual cost of unwanted renewals: EUR 3.2 million. A mid-market PE acquirer used Kira on an 8,000-contract data room and identified 340 change-of-control termination rights that reshaped the deal terms. These are the use cases that justify the investment.

But I have also seen the failure mode. A company spent six figures on a platform, loaded their contracts, and discovered that 40% existed only as scanned PDFs of varying quality. OCR errors compounded extraction errors. Handwritten amendments were illegible. The platform showed "low confidence" on roughly 2,000 contracts, all of which still required manual review. The promise of portfolio intelligence ran headfirst into the reality of document quality. Nobody had budgeted for the remediation.

The European complexity multiplier. Cross-border European operations add layers that American-focused platforms were not designed for. A German manufacturer with Polish suppliers, Czech joint ventures, and French distribution agreements needs contract intelligence that handles multiple languages, legal systems, and contractual conventions simultaneously. Most platforms perform well on English-language common law contracts. Performance degrades on civil law instruments, particularly those drafted in languages with smaller training corpora. Before purchasing any platform for European operations, test it on your actual multilingual contract portfolio — not the vendor's English-language demo.

The real differentiator is no longer review speed. Everyone can do that. It is what you do with the intelligence you extract. And that requires asking the hard question before signing: do we have the document quality, the defined use case, and the operational commitment to make this work? If you cannot define that use case for your organisation, the platform is an expense, not an investment.

---

The Competence Question

Imagine you are advising a mid-market manufacturer whose procurement team just signed a three-year contract intelligence platform licence. They ask you to review the vendor agreement. The platform's terms of service grant the vendor broad rights to use uploaded contract data for model training. Your client's contracts contain customer pricing, supplier terms, and competitive intelligence.

Here is the competence problem: most lawyers reviewing this vendor agreement would focus on standard commercial terms — limitation of liability, indemnification, termination. But the meaningful risk is in the data handling provisions, and evaluating those provisions requires understanding how large language models actually process and retain training data. Do you know whether the vendor's "anonymisation" claims hold up technically? Can you advise your client on whether their trade secrets are genuinely protected?

Contract intelligence platforms are not just legal tools. They are data infrastructure. And advising clients on their procurement requires a level of technical understanding that most commercial lawyers have not yet developed.

I encountered a variation of this last year when a client's IT team pushed back on their legal department's platform choice, arguing the vendor's data retention practices were incompatible with their information security policy. The lawyers had not consulted IT before signing. The lawyers had not read the technical documentation. The deal had to be unwound at significant cost and embarrassment.

For European firms, the GDPR dimension compounds this. Contract data uploaded to platforms hosted outside the EU triggers data transfer obligations. The AI Act adds further requirements around transparency and documentation for AI systems processing business data. The question is not whether you use these platforms yourself. It is whether you can competently advise clients who do — across the full range of commercial, data protection, and AI regulatory obligations.

---

What To Do

1. Define your question before shopping. Audit your contract workflows and identify the specific intelligence gap — auto-renewal exposure, clause deviation patterns, portfolio risk concentration. If you cannot name the problem, you are not ready for the platform.

2. Test with your contracts, not demo data. Every platform performs well on clean, standard agreements. Request a proof of concept using your actual documents, including legacy scanned PDFs, non-standard clause language, and — for European operations — multilingual contracts. Accuracy drops significantly on messy inputs.

3. Read the data handling terms. Before uploading your contract portfolio to any platform, understand exactly how inputs are processed, retained, and potentially used for training. For European operations, assess GDPR compliance and cross-border data transfer implications. This applies to your own use and to advising clients.

4. Budget for the full cost. Platform licensing is typically 40-60% of total cost. Add implementation, legacy document remediation, custom model training, and ongoing IT partnership. A EUR 100K licence can easily become a EUR 200K commitment.

5. Watch the consolidation. Before committing to a multi-year agreement, assess the vendor's independence. Acquired platforms change roadmaps. Ask explicitly about product continuity commitments. Evisort's post-Workday evolution is a cautionary example.

---

Quick Reads

• Ivo raises $55M Series B at $355M valuation, targeting in-house legal teams with contract intelligence.

• Workday's Evisort acquisition signals enterprise software companies absorbing legal AI rather than partnering with it.

• Kira reports 450,000+ documents processed monthly, maintaining its position as the M&A due diligence standard.

• The contract intelligence market is projected to grow at 23% CAGR, reaching $7.2B by 2033 — but the mid-market accessibility gap remains the sector's unresolved problem.

---

One Question

If contract intelligence platforms can surface risks across your entire portfolio that no human review ever could, what happens to the standard of care for lawyers who choose not to use them?

---

TwinLadder Weekly | Issue #17 | October 2025

Compliance is the floor. Competence is the mission.