Litera Review for Lawyers: Document Comparison, Drafting, and the Confidentiality Questions Worth Asking
Litera is one of the most entrenched names in legal drafting technology. If you have ever compared two versions of an agreement in Word, chances are you have used Litera Compare (formerly Workshare Compare) or its ancestor DeltaView. Over the last few years, Litera has expanded well beyond redlining into a broader drafting, proofreading, transaction management, and generative-AI stack. This review looks at Litera through the lens a practicing lawyer actually cares about: what it does well, where client data lives, how it maps to ABA Formal Opinion 512, and whether it justifies its price for your practice.
This review focuses on the drafting and comparison side of Litera's portfolio — Compare, Check, Draft, and the newer AI features layered on top — rather than the transaction management or DMS integrations, which deserve their own treatment.
What Litera Actually Does
Litera is a suite, not a single product. The pieces most drafting lawyers touch are:
- Litera Compare — the document comparison engine that produces redlines between two versions of a Word, PDF, or scanned document. It handles heavy formatting, tracked changes, and cross-format comparison (e.g., PDF vs. Word) better than Word's native comparison, which is the reason most firms bought it in the first place. See the Litera Compare product page.
- Litera Check — a proofreading and contract-integrity tool that flags defined-term issues, inconsistent numbering, broken cross-references, phone-number and date errors, and similar drafting bugs that human eyes miss at 11pm.
- Litera Draft — a drafting assistant that surfaces clauses and templates and integrates with the firm's knowledge base.
- Generative AI features — Litera has been rolling out AI-assisted summarization, clause analysis, and drafting suggestions under the broader Litera One umbrella. Current capabilities and the underlying model providers are described on the Litera products page.
The core comparison engine is the crown jewel. It is fast, the redlines are clean, and it survives messy real-world documents (opposing counsel's PDF with track changes accepted, a scanned amendment, a Word file that has been round-tripped through three drafting systems). For transactional lawyers, that alone is often the sale.
The Confidentiality Lens
This is the section that matters most, and the answer with Litera is genuinely nuanced because different products in the suite have different data paths.
Litera Compare (desktop) historically ran entirely on the local machine — you fed it two documents, it produced a redline, nothing traversed the internet. The desktop comparison product remains a local-processing tool for the core comparison operation, which is one reason it has been so widely accepted by risk-averse firms. Confirm your specific deployment with your IT team, because Litera also offers cloud-hosted and integrated versions where processing happens on Litera's infrastructure.
Cloud and AI-assisted features are a different story. When you use Litera's generative AI features, or any of the cloud-hosted collaboration and transaction products, document content is transmitted to Litera's cloud environment and, in some configurations, to third-party model providers Litera has contracted with. Litera publishes a Trust Center with security documentation, subprocessor lists, and certifications (SOC 2, ISO 27001) that you should read before enabling AI features firm-wide.
Key questions to answer with your Litera account team, in writing, before rollout:
- For each product you deploy, does the document content leave the endpoint, and if so, to which cloud regions and which subprocessors?
- Is client content used to train or fine-tune any models — Litera's own, or a vendor's? Get the "no training" commitment in the order form or DPA, not just marketing copy.
- What is the retention period for prompts, outputs, and any intermediate content on Litera's systems?
- Will Litera sign a Data Processing Addendum, and, if you handle protected health information, a Business Associate Agreement?
- What would a third-party subpoena served on Litera actually reach — original documents, redlines, prompts, outputs, metadata, logs? Litera should be able to describe this, and the answer should match its contractual retention commitments.
The general architectural point: local desktop comparison has a very small confidentiality surface. Cloud drafting and AI features have a larger one. That is not a reason to avoid them — it is a reason to configure them deliberately.
Ethics-Opinion Fit (ABA 512)
ABA Formal Opinion 512 (July 2024) is now the reference point for lawyers using generative AI. It frames the analysis around competence (Model Rule 1.1), confidentiality (Model Rule 1.6), communication with clients, supervisory responsibility, and reasonable fees. Applied to Litera:
- Competence. Compare and Check are mature deterministic tools. The output is deterministic redlines and rule-based error flags, and lawyers already know how to supervise that. The generative AI features require the same skepticism you would bring to any large-language-model output — hallucinated clauses, subtly wrong summaries, plausible-sounding but incorrect defined-term interpretations. Read outputs; do not paste them.
- Confidentiality. ABA 512 emphasizes understanding where inputs go and whether they train models. That drives directly to the DPA and subprocessor questions above. NYC Bar Formal Opinion 2025-6 (December 2025) reinforces the same points at the state level, with particular attention to informed client consent when using self-learning tools.
- Client communication. Whether you need to tell clients you use Litera depends on engagement terms and what features you enable. For deterministic redlining, most firms treat it like Word — no separate disclosure. For generative drafting on client documents, many engagement letters now include AI-use language.
- Waiver risk. The relevant recent decisions — US v. Heppner (S.D.N.Y. Feb 2026), where Judge Rakoff held that a litigant's chats with a public AI platform were not privileged, and Brewer v. Otter.ai (2025) on notetaker capture — both concern consumer AI tools without enterprise contractual protections. Litera's enterprise contract path is materially different, but the underlying lesson (know exactly who has your data) still applies.
Where Litera Is Strong
Three things Litera does well that competitors still struggle with:
- Comparison quality on ugly documents. Cross-format comparison, tolerance for scanned inputs, and clean handling of tables and numbered lists remains best-in-class. This is decades of accumulated engineering, not a feature a startup can replicate over a weekend.
- Word integration. Litera lives inside Word, which is where transactional drafting actually happens. Ribbon buttons, keyboard shortcuts, and iManage/NetDocuments integration mean it slots into existing workflows.
- Enterprise contracting posture. Litera is used to negotiating with law firm risk and IT teams. DPAs, security questionnaires, and audit rights are routine, not exceptional requests.
Where to Be Careful
- Feature sprawl. The Litera catalog is large, and pricing and functionality vary across bundles. Buy for the two or three products you will actually use.
- AI feature maturity. The generative AI features are newer than the deterministic tools. Pilot them on non-sensitive matters, and measure error rates before firm-wide rollout.
- Deployment mode matters. The confidentiality profile of "Litera Compare desktop" is not the same as "Litera in the cloud with AI enabled." Do not let a marketing conversation blur those.
- Word-centric. If your drafting workflow is heavily Google Docs, Litera is less native than it is in a Microsoft shop.
Pricing and Who It's For
Litera does not publish list pricing publicly. Firms typically license per user, with volume discounts and bundle pricing, and negotiate multi-year enterprise agreements. Rather than quote numbers that may be stale by the time you read this, ask for a quote through the Litera contact page and benchmark it against at least one competing comparison tool.
Who should buy Litera:
- Transactional practices where document comparison is a daily activity — M&A, finance, real estate, commercial contracting.
- Firms already standardized on iManage or NetDocuments, where Litera's integrations reduce friction.
- Litigation groups that regularly compare marked-up pleadings, exhibits, or produced documents across formats.
Who probably does not need the full suite:
- Solos and small firms whose comparison needs are met by Word's native tool and who cannot amortize an enterprise contract across enough matters.
- Practices that live in Google Docs and rarely round-trip to Word.
How Litera Fits Alongside Other Tools
Litera is a drafting and comparison tool, not a meeting capture tool, so it does not overlap with notetaker products. Firms that use Litera for drafting often pair it with a separate, on-device notetaker for client calls and internal strategy meetings — see Basil for Law for how a local-only capture tool complements a cloud drafting stack without adding a second cloud data path. Read the two tools' data-handling terms together so you understand the full picture across your workflow.
Verdict
Litera Compare is a reference-quality tool in its category, and the broader drafting suite is a credible enterprise offering. The AI-assisted features are worth piloting but should be procured with the same rigor as any other cloud AI product: DPA, subprocessor list, training-use commitment, retention terms, and a clear-eyed view of subpoena exposure. For transactional firms, Litera is often the right answer; the discipline is in scoping the deployment and negotiating the paper.
| Pros | Cons |
|---|---|
| Best-in-class document comparison, including cross-format | Feature and pricing complexity across the suite |
| Deep Word and DMS integration | AI features newer and require pilot testing |
| Desktop Compare has a small confidentiality surface | Cloud and AI features expand the data footprint materially |
| Enterprise-grade contracting, DPA and security documentation available | List pricing not public; requires negotiation |
| Mature user base and training materials | Less native for Google Docs-first practices |
This review is for information only and is not legal advice.
Frequently asked questions
Does Litera Compare send my documents to the cloud?
The traditional Litera Compare desktop product performs the core comparison locally on the user's machine. Cloud-hosted variants, AI-assisted features, and certain integrations do transmit content to Litera's infrastructure. Confirm the exact data path for your deployment with Litera and your IT team, and review the Litera Trust Center documentation before enabling AI features.
Is Litera a good fit under ABA Formal Opinion 512?
For deterministic tools like Compare and Check, the analysis is similar to any established legal software — supervise the output and understand the data path. For generative AI features, ABA 512 expects you to understand where inputs go, whether they train models, and how outputs are retained. Get those commitments in the DPA, not just marketing materials.
Will Litera sign a DPA or BAA?
Litera is accustomed to enterprise contracting with law firms and generally offers a Data Processing Addendum. If you handle protected health information, ask specifically about a Business Associate Agreement, as availability may depend on the products in scope. Request the current templates through your account team.
How does Litera compare to Word's native document comparison?
Word's native comparison is adequate for simple Word-to-Word redlines between clean documents. Litera Compare handles cross-format comparison (PDF vs. Word, scanned documents), heavy formatting, tables, and messy real-world inputs more reliably, which is why most firms with transactional volume standardize on it.
What does Litera cost?
Litera does not publish list pricing. Licensing is typically per user with volume and multi-year discounts, and pricing varies by which products in the suite you deploy. Request a quote through Litera's contact page and benchmark against at least one competing comparison or drafting tool before signing.
Do I need to tell clients I use Litera?
For deterministic redlining most firms treat Litera like Word and do not disclose separately. For generative AI features applied to client documents, many engagement letters now include general AI-use language. Check your jurisdiction's guidance, including ABA Formal Opinion 512 and NYC Bar Formal Opinion 2025-6, and align your engagement terms accordingly.
Meeting notes with no server to subpoena
Basil transcribes and summarizes entirely on-device — privilege-safe by architecture. See Basil for Law →
This review is for information only and is not legal advice.