# Mage — AI M&A Legal Diligence

Mage is an AI platform built for M&A diligence. We ingest data rooms, run risk-driven document review, resolve amendment chains, draft memos and disclosure schedules, and review counterparty redlines. Built by an ex-Kirkland & Ellis M&A attorney and an ex-Google Cloud Document AI engineer.

Curated library below. Pillar guides first, then flagship posts (sorted by traffic).

## Pillar guides

- [Evaluating Legal AI Tools: A Buyer's Guide for M&A Counsel](https://magelegal.com/guide/evaluating-legal-ai-tools): How M&A counsel should actually evaluate legal AI tools before deploying on real deals. Accuracy methodology, security posture, workflow fit, and the questions vendors hate. _(author: Mage Team; 1693 words; written)_
- [Legal AI vs. Harvey vs. Generic AI: How to Evaluate](https://magelegal.com/guide/legal-ai-vs-harvey-vs-generic): An honest comparison of legal-AI categories: Harvey-style firm-wide assistants, generic LLMs (ChatGPT, Claude, Gemini), and M&A-specific tools like Mage. What each is good for, where each fails. _(author: Mage Team; 1353 words; written)_
- [Legal AI for M&A: The Practitioner's Guide](https://magelegal.com/guide/legal-ai-for-ma): How AI is changing M&A diligence, drafting, and closing for buy-side and sell-side counsel. The categories that matter, the failure modes to watch for, and what to evaluate. _(author: Mage Team; 2118 words; written)_
- [AI Due Diligence: An Operational Playbook](https://magelegal.com/guide/ai-due-diligence): How to run AI-augmented buy-side and sell-side diligence on a real deal. Step-by-step from data room access to signed memo, with the failure modes to watch for. _(author: Mage Team; 1708 words; written)_

## Flagship posts

- [The F1 Engine Problem: Why AI Disappointment Has Nothing to Do with AI](https://magelegal.com/blog/f1-engine-problem): Every legal team has the same engine. Most strapped it to a bicycle. The disappointment with AI isn't the model — it's everything around the model. _(author: Raffi Isanians; 1700 words; written)_
- [Most 'Legal AI' Is Just a Foundation Model Behind a Brand. Here's Why That's Not Enough.](https://magelegal.com/blog/legal-native-intelligence): Foundation models extract text. Legal-Native Intelligence understands how legal documents actually work. Here's what that means and why it matters for M&A diligence. _(author: Raffi Isanians; 1381 words; written)_
- [Everything You Need to Know About Prompting AI You Learned in Law School](https://magelegal.com/blog/everything-you-need-to-know-about-prompting-ai-you-learned-in-law-school): The most effective AI prompting techniques, specificity, structured frameworks, strategic context, targeted follow-up, are skills every lawyer learned in law school. Research proves it. _(author: Raffi Isanians; 3643 words; written)_
- [LLM Hallucination in Contract Analysis: Why Source Verification Is Non-Negotiable](https://magelegal.com/blog/llm-hallucination-in-contract-analysis): Large language models hallucinate. In legal contract analysis, a single fabricated clause citation can derail a deal. Here is how hallucination manifests in legal AI, why it happens, and how to build systems that prevent it. _(author: Raffi Isanians; 1561 words; written)_
- [Signing to Closing: Interim Covenants and Compliance Monitoring in M&A](https://magelegal.com/blog/signing-to-closing-interim-covenants): A practical guide to managing the period between signing and closing in M&A transactions. Covers interim operating covenants, compliance monitoring, material adverse change triggers, and how to avoid deal-threatening breaches. _(author: Mage Team; 1696 words; written)_
- [What 300 NDAs Taught Me About Change of Control Clauses](https://magelegal.com/blog/what-300-ndas-taught-me-about-change-of-control-clauses): After analyzing change of control provisions across 300 NDAs for M&A transactions, clear patterns emerge. Five categories of COC provisions, each with different implications for deal execution and post-closing risk. _(author: Raffi Isanians; 1627 words; written)_
- [Non-Compete Clauses in M&A: Enforceability, Extraction, and Deal Impact](https://magelegal.com/blog/non-compete-clauses-in-manda-enforceability-extraction-deal-impact): Non-compete clauses are among the most scrutinized provisions in M&A transactions. This guide covers the evolving enforceability landscape, how geographic and temporal scope affect deal value, and why systematic extraction across a data room matters. _(author: Mage Team; 1608 words; written)_
- [Exclusivity Clauses in Commercial Contracts: What M&A Deal Teams Need to Know](https://magelegal.com/blog/exclusivity-clauses-in-commercial-contracts): Exclusivity clauses in a target's commercial contracts can fundamentally reshape post-acquisition strategy. This guide covers the types of exclusivity provisions, how they affect deal value, and why they require early identification during diligence. _(author: Mage Team; 1544 words; written)_
- [Anti-Assignment Clauses in M&A: What Every Deal Attorney Should Know](https://magelegal.com/blog/anti-assignment-clauses-in-manda-what-every-deal-attorney-should-know): Anti-assignment clauses can derail acquisitions when they go undetected in a data room. This guide covers the types of anti-assignment provisions, how they interact with change of control transactions, and what deal teams should flag during diligence. _(author: Mage Team; 1542 words; written)_
- [Most Favored Nation Clauses in M&A: Pricing, Compliance, and Deal Impact](https://magelegal.com/blog/most-favored-nation-clauses-mfn-manda): Most favored nation clauses obligate the target to offer its best pricing or terms to specific counterparties. This guide covers how MFN provisions work, why they matter in acquisitions, and how they create ongoing compliance obligations post-closing. _(author: Mage Team; 1530 words; written)_
- [LOI to Closing: The Deal Attorney's Diligence Timeline](https://magelegal.com/blog/loi-to-closing-deal-attorney-diligence-guide): A practical timeline for M&A due diligence from letter of intent through closing. Covers critical path items, milestone sequencing, and how to avoid the bottlenecks that delay deals. _(author: Mage Team; 1455 words; written)_
- [Financial Services M&A: Regulatory Approvals and Compliance Due Diligence](https://magelegal.com/blog/financial-services-manda-regulatory-approval): A practical guide to financial services M&A due diligence for deal attorneys. Covers FDIC and OCC approval processes, state banking regulator requirements, broker-dealer compliance, and the regulatory complexities that extend deal timelines. _(author: Mage Team; 1789 words; written)_
- [Manufacturing M&A Due Diligence: Environmental Liabilities and Supply Chain Risks](https://magelegal.com/blog/manufacturing-manda-environmental-supply-chain): A practical guide to manufacturing M&A due diligence for deal attorneys. Covers environmental liability assessment, supply chain contract review, equipment leases, union agreements, and the operational risks unique to manufacturing targets. _(author: Mage Team; 1705 words; written)_
- [Why We Do Not Let Users Write Prompts](https://magelegal.com/blog/why-we-dont-let-users-write-prompts): Open prompt boxes in legal AI create security risks, accuracy problems, and inconsistent output. Constrained interfaces that encode domain expertise produce better results for M&A attorneys than flexible prompt engineering ever will. _(author: Raffi Isanians; 1427 words; written)_
- [How to Build a Due Diligence Checklist for M&A](https://magelegal.com/blog/how-to-build-due-diligence-checklist-manda): A comprehensive guide to building an M&A due diligence checklist tailored to your deal type. Covers the essential categories, how to customize for buy-side vs. sell-side transactions, and what experienced deal teams include that checklists from templates miss. _(author: Mage Team; 1765 words; written)_
- [Model Fusion: Why a Single AI Model Is Not Enough for Legal Document Analysis](https://magelegal.com/blog/model-fusion-technology-why-single-model-not-enough): Why Mage uses multiple specialized AI models instead of relying on a single general-purpose model. Covers the limitations of single-model approaches, how model fusion works, and why ensemble methods deliver the precision that legal work demands. _(author: Raffi Isanians; 1737 words; written)_
- [Technology M&A Due Diligence: Software Licenses, IP Chains, and Data Privacy](https://magelegal.com/blog/technology-manda-software-license-ip-diligence): A practical guide to technology M&A due diligence for deal attorneys. Covers software license review, SaaS agreement analysis, IP assignment chain verification, open source compliance, and data privacy assessment. _(author: Mage Team; 1661 words; written)_
- [Governing Law and Forum Selection Clauses in M&A: Why Jurisdiction Matters](https://magelegal.com/blog/governing-law-forum-selection-manda): Governing law and forum selection clauses determine where disputes are resolved and which law applies. This guide covers why these provisions matter in M&A, how multi-jurisdiction complications arise, and what deal teams should extract during diligence. _(author: Mage Team; 1630 words; written)_
- [The Real Bottleneck in M&A Diligence Isn't the Documents. It's the Workflow.](https://magelegal.com/blog/real-bottleneck-manda-diligence-workflow): Most legal teams lose days not because they lack information, but because they lack a system for processing it. Here's how AI-powered document review is changing that. _(author: Raffi Isanians; 1628 words; written)_
- [Termination for Convenience vs. Cause: What M&A Attorneys Must Know](https://magelegal.com/blog/termination-for-convenience-vs-cause): Termination provisions determine the stability of every contract in a target's portfolio. This guide explains the difference between termination for convenience and termination for cause, how notice and cure rights work, and why these clauses are critical for M&A diligence. _(author: Mage Team; 1592 words; written)_
- [Harvey vs. Kira vs. Infrastructure: Three Approaches to Legal AI](https://magelegal.com/blog/harvey-vs-kira-vs-infrastructure-legal-ai): The legal AI market has consolidated around three paradigms: research assistants (Harvey), legacy extraction platforms (Kira), and purpose-built infrastructure (Mage). They solve different problems for different workflows. _(author: Raffi Isanians; 1575 words; written)_
- [Why Associates Spend 60 Hours on Material Contracts (and How to Reclaim It)](https://magelegal.com/blog/why-associates-spend-60-hours-on-material-contracts): The material contracts review in an M&A deal breaks down into four phases: reading (30 hours), flagging (15 hours), summarizing (10 hours), and formatting (5 hours). Most of that time is information processing, not legal analysis. _(author: Raffi Isanians; 1514 words; written)_
- [Why Clause-Level Segmentation Changes Everything in Legal AI](https://magelegal.com/blog/clause-level-segmentation-precision): Most legal AI tools analyze documents at the page or paragraph level. Clause-level segmentation, where the system understands individual provisions as discrete units, is the difference between approximate summaries and precise, verifiable extraction. _(author: Raffi Isanians; 1446 words; written)_
- [How Buy-Side Team Used Mage for Post-Signing Covenant Compliance](https://magelegal.com/blog/how-buy-side-team-used-mage-for-post-signing-covenant-compliance): Case study: A buyer uses AI-powered contract analysis to monitor seller compliance with operating covenants during the 90-day period between signing and closing. _(author: Mage Team; 1301 words; written)_
- [M&A Trends by Industry: A Practitioner's Survey](https://magelegal.com/blog/ma-trends-by-industry): How M&A diligence diverges by sector. Working capital quirks, regulatory traps, IP exposure, and integration realities across 12 industries M&A counsel actually transact in. _(author: Mage Team; 3400 words; written)_
- [How to Roll Out Legal AI at a Law Firm](https://magelegal.com/blog/how-to-roll-out-legal-ai-at-a-law-firm): A 90-day rollout playbook for legal AI inside an M&A practice. How to pilot, when to standardize, how to handle change-management without losing junior morale. _(author: Raffi Isanians; 2700 words; written)_
- [Amendment Chain Resolution: The Hardest Problem in Legal AI](https://magelegal.com/blog/amendment-chain-resolution-hardest-problem-legal-ai): Why amendment chains break standard AI document analysis approaches, how structured extraction handles them, and what makes multi-amendment resolution the defining technical challenge for legal AI systems. _(author: Raffi Isanians; 1977 words; written)_
- [How to Write a Due Diligence Memo for M&A](https://magelegal.com/blog/how-to-write-due-diligence-memo): The due diligence memo is the primary deliverable that translates contract review into deal intelligence. This guide covers the structure, content standards, and practical techniques for writing diligence memos that partners and clients actually use. _(author: Mage Team; 1926 words; written)_
- [Why We Built a Legal Document Classifier First](https://magelegal.com/blog/why-we-built-legal-document-classifier-first): Why Mage built document classification before extraction, how document types determine extraction strategy, and why getting classification right is the prerequisite for everything else in legal AI. _(author: Raffi Isanians; 1904 words; written)_
- [How to Architect a Document AI Pipeline for Legal](https://magelegal.com/blog/how-to-architect-a-document-ai-pipeline-for-legal): A practitioner's view of the architecture under modern legal AI: ingestion, classification, structured extraction, multi-document reasoning, validation, and output. What goes where, and why. _(author: Raffi Isanians; 1850 words; written)_
- [What Is Legal AI, Really?](https://magelegal.com/blog/what-is-legal-ai-really): A direct answer for attorneys searching the question. The category, the categories of tool inside it, what each does well, and where each falls short — written for a partner deciding whether to deploy. _(author: Mage Team; 1800 words; written)_
- [Why Most Legal AI Fails: Three Failure Modes That Kill Adoption](https://magelegal.com/blog/why-most-legal-ai-fails-kirkland-perspective): After speaking with hundreds of M&A attorneys, three patterns explain why most legal AI tools fail to achieve adoption: wrong abstraction level, no workflow integration, and output below the trust threshold. _(author: Raffi Isanians; 1793 words; written)_
- [How to Evaluate Legal AI Tools for M&A: A 5-Axis Framework](https://magelegal.com/blog/legal-ai-tools-for-manda-evaluation-framework): Not all legal AI tools solve the same problem. Here is a framework for evaluating them across the five dimensions that actually matter for M&A deal teams: accuracy, speed, security, setup cost, and output quality. _(author: Mage Team; 1760 words; written)_
- [How to Review 500 Contracts in a Weekend](https://magelegal.com/blog/how-to-review-500-contracts-in-a-weekend): A step-by-step guide for M&A attorneys facing a tight diligence timeline. This post covers the triage, prioritization, extraction, and review workflow that makes high-volume contract review manageable, with and without AI assistance. _(author: Mage Team; 1745 words; written)_
- [Indemnification Caps and Baskets: How They Shape M&A Deal Economics](https://magelegal.com/blog/indemnification-caps-and-baskets-deal-economics): Indemnification caps and baskets are among the most negotiated provisions in M&A purchase agreements. This guide explains how deductible and tipping baskets work, current market standards, and how these mechanisms allocate post-closing risk between buyer and seller. _(author: Mage Team; 1736 words; written)_
- [How to Identify Material Contracts in a Data Room](https://magelegal.com/blog/how-to-identify-material-contracts-data-room): Identifying which contracts are material is the first and most important judgment call in M&A due diligence. This guide covers value thresholds, strategic importance criteria, unusual terms flags, and SEC materiality standards that experienced deal teams use to separate the critical from the routine. _(author: Mage Team; 1724 words; written)_
- [Tech M&A Diligence: What Software Targets Actually Need](https://magelegal.com/blog/tech-ma-diligence-software-targets): Tech M&A is intangibles-first. The valuation rests on IP integrity, customer concentration, and open-source compliance. A practitioner's view of what to check on a software target and where deals actually break. _(author: Mage Team; 1700 words; written)_
- [How Attorneys Should Evaluate LLM-Powered Tools](https://magelegal.com/blog/how-attorneys-should-evaluate-llm-powered-tools): An evaluator's framework for attorneys reviewing legal AI tools. The questions vendors hate, the metrics that matter, and the procurement traps that cost firms months. _(author: Mage Team; 1700 words; written)_
- [Mage vs. Harvey: A Feature-by-Feature Comparison for M&A Counsel](https://magelegal.com/blog/mage-vs-harvey-feature-by-feature): An honest, sourced comparison of Mage and Harvey for M&A diligence work. Where each is built to win, where each falls short, and how to evaluate them on a real deal. _(author: Mage Team; 1700 words; written)_
- [Research vs. Extraction: Two Paradigms for Contract Review Software](https://magelegal.com/blog/contract-review-software-comparison-research-vs-extraction): Contract review software falls into two paradigms: research tools that answer questions about documents, and extraction tools that systematically pull structured data from every contract. The distinction determines what you can build on top of the output. _(author: Raffi Isanians; 1691 words; written)_
- [IP Assignment Clauses PE Buyers Miss: Chain of Title Risks in M&A](https://magelegal.com/blog/ip-assignment-clauses-pe-buyers-miss): Gaps in IP assignment clauses are among the most overlooked risks in private equity acquisitions. This guide covers chain of title issues, employee invention assignments, contractor work-for-hire provisions, and open source risks that PE deal teams should flag during diligence. _(author: Mage Team; 1664 words; written)_
- [How We Test Legal AI Accuracy: Mage's Benchmarking Methodology](https://magelegal.com/blog/how-we-test-legal-ai-accuracy-benchmarking): An inside look at how Mage benchmarks the accuracy of its legal AI system. Covers test methodology, human reviewer comparison, confidence scoring, and why accuracy without a rigorous testing framework is just a marketing number. _(author: Raffi Isanians; 1640 words; written)_
- [Post-Closing Integration: How to Hand Off Diligence Findings Without Losing Institutional Knowledge](https://magelegal.com/blog/post-closing-integration-diligence-handoff): A practical guide to transitioning from M&A due diligence to post-closing integration. Covers TSA setup, consent tracking, compliance calendars, and how to prevent the institutional knowledge loss that derails integrations. _(author: Mage Team; 1598 words; written)_
- [Why We Built Mage After Kirkland](https://magelegal.com/blog/why-we-built-mage-after-kirkland): I spent years inside one of the most demanding M&A practices in the world. The bottleneck wasn't the work — it was the time spent doing the wrong parts of it. That's why Mage exists. _(author: Raffi Isanians; 1500 words; written)_
- [PE Add-On Diligence: An Integrated Playbook](https://magelegal.com/blog/pe-add-on-diligence-playbook): Add-on diligence is platform-thesis-first. The diligence question is whether the target plugs into the platform without breaking it. A playbook for PE counsel on what to check and how to compress the timeline. _(author: Mage Team; 1500 words; written)_
- [SOC 2 and Legal AI: What M&A Lawyers Should Demand](https://magelegal.com/blog/soc2-and-legal-ai-what-ma-lawyers-should-demand): SOC 2 Type II is the bar, not the goal. The questions M&A lawyers should be asking legal AI vendors before privileged content touches the platform — and the deflections to watch for. _(author: Mage Team; 1500 words; written)_
- [Disclosure Schedule Traps Sellers' Counsel Misses](https://magelegal.com/blog/disclosure-schedule-traps-sellers-counsel-misses): The disclosure schedule is where post-closing indemnification claims are won and lost. The recurring traps that cost sellers money, and how to avoid them. _(author: Mage Team; 1500 words; written)_
- [Mage vs. Kira: How They Compare for M&A Diligence](https://magelegal.com/blog/mage-vs-kira-feature-by-feature): Mage and Kira occupy different generations of the contract analysis category. An honest comparison for M&A counsel deciding between extraction-first and workflow-first. _(author: Mage Team; 1500 words; written)_
- [The ROI of Legal AI for M&A: An Attorney's Calculator](https://magelegal.com/blog/roi-of-legal-ai-for-ma): Concrete numbers for evaluating legal AI ROI on M&A practice: hours saved, cost-per-deal, output quality. Built from real customer deployments and the math attorneys actually use. _(author: Mage Team; 1500 words; written)_
- [How a Mid-Market Firm Reviewed 2,000 Contracts in 48 Hours](https://magelegal.com/blog/how-mid-market-firm-reviewed-2000-contracts-in-48-hours): Case study: A mid-market law firm uses AI to complete first-pass review of a 2,000-document data room in 48 hours, identifying critical change of control issues before the buyer walked. _(author: Mage Team; 1462 words; written)_
- [How a PE Buyer Identified $5M in Hidden Liabilities](https://magelegal.com/blog/how-pe-buyer-identified-$5m-in-hidden-liabilities): Case study: A private equity fund uses AI-powered due diligence to uncover $5M in undisclosed liability exposure buried across 400+ vendor contracts. _(author: Mage Team; 1406 words; written)_
- [How a Law Firm Reduced Diligence Costs by 60%](https://magelegal.com/blog/how-law-firm-reduced-diligence-costs-by-60%): Case study: A mid-market law firm implements AI-powered due diligence across its M&A practice, reducing first-pass review costs by 60% while improving quality. _(author: Mage Team; 1405 words; written)_
- [How a Seller's Counsel Built Disclosure Schedules 10x Faster](https://magelegal.com/blog/how-sellers-counsel-built-disclosure-schedules-10x-faster): Case study: Seller's counsel uses AI to transform a 3-week disclosure schedule project into 3 days, catching issues before buyer diligence began. _(author: Mage Team; 1381 words; written)_
- [Mage vs. Luminance: How They Compare for M&A Diligence](https://magelegal.com/blog/mage-vs-luminance-feature-by-feature): Mage and Luminance both serve transactional teams with AI contract review. An honest comparison for M&A counsel — architecture, workflow, and where each is built to win. _(author: Mage Team; 1350 words; written)_
- [Mage vs. ContractPodAi: How They Compare for M&A Counsel](https://magelegal.com/blog/mage-vs-contractpodai-feature-by-feature): Mage and ContractPodAi solve different problems. ContractPodAi is contract lifecycle management; Mage is M&A diligence. Why this matters when choosing for an M&A practice. _(author: Mage Team; 1331 words; written)_
- [What I Got Wrong About Legal AI](https://magelegal.com/blog/what-i-got-wrong-about-legal-ai): Three predictions I made about legal AI in 2023 that turned out to be wrong, what I learned from the misses, and what I think now. _(author: Raffi Isanians; 1300 words; written)_
- [Mage vs. Legora: How They Compare for M&A Counsel](https://magelegal.com/blog/mage-vs-legora-feature-by-feature): Mage and Legora are both modern, LLM-native legal AI platforms but built for different scopes. An honest comparison for M&A practices choosing between firm-wide and specialist. _(author: Mage Team; 1300 words; written)_
- [Mage Product Walkthrough: What the Tool Actually Does](https://magelegal.com/blog/mage-product-walkthrough): A workflow walkthrough of what Mage does on a real deal. From data room access on Day 1 to partner-reviewable memo on Day 4. Honest, not a marketing tour. _(author: Mage Team; 1300 words; written)_
- [Multi-Document Context Windows in Legal AI, Explained](https://magelegal.com/blog/multi-document-context-windows-explained): Why bigger model context windows don't solve multi-document legal reasoning. The architectural problems context windows can't paper over, and what works instead. _(author: Raffi Isanians; 1300 words; written)_

## Where to start

- Master hub: https://magelegal.com/guide/legal-ai-for-ma
- About: https://magelegal.com/about
- Team: https://magelegal.com/team
- Security: https://magelegal.com/security
- Request a demo: https://magelegal.com/request-demo