> ## Content Index > Fetch the complete content index at: https://adikumar.co/llms.txt > Use this file to discover other available public pages before exploring further. # Legal and contract DD - URL: https://adikumar.co/legal-and-contract-dd/ - Published: 2026-08-28T18:42:02.000Z - Updated: 2026-10-07T05:39:04.000Z - Description: Corporate structure, material contracts, change-of-control provisions, non-compete enforceability, IP filings and freedom-to-operate, litigation exposure. Where the deal thesis meets the paper record. - Author: Adi Kumar - Tags: Due Diligence for the AI Buildout, AI Infrastructure, M&A, #Essay [Due Diligence for the AI Buildout](https://adikumar.co/due-diligence-for-the-ai-buildout/) · Part 10 of 14 · Structural + paper · [Framework page](https://adikumar.co/framework-due-diligence/) Due Diligence for the AI Buildout · Part X of XIV # Legal and contract DD The workstream where the deal thesis meets the actual paper record, from corporate structure and material contracts through change-of-control provisions, IP filings, litigation exposure and employment law. Executed almost entirely by external counsel. Due Diligence for the AI Buildout · fourteen essays for practitioners 1. **Part I.** The scoping decision 2. **Part II.** Team, timeline, data room 3. **Part III.** Market and demand-side DD 4. **Part IV.** Customer, pricing, competitive DD 5. **Part V.** Product architecture and technology DD 6. **Part VI.** Manufacturing, IP, supply chain DD 7. **Part VII.** Quality of earnings and working capital DD 8. **Part VIII.** Capex, growth economics, stress-testing 9. **Part IX.** Organisational, key personnel, culture, systems 10. **Part X.** Legal and contract DD *(you are here)* 11. **Part XI.** Regulatory, CFIUS/HSR, environmental, cyber 12. **Part XII.** Deal structure, reps and warranties, escrow 13. **Part XIII.** Financing structure, covenants, closing risk 14. **Part XIV.** Post-close validation and the DD improvement loop Supplements 1. **A.** The Question Banks (downloadable) 2. **B.** Worked example: Eaton / Boyd Thermal end-to-end 3. **C.** Seller-side prep: the DD workbook you should have ready #### Executive summary - Legal DD covers six territories: corporate structure and cap table, customer contracts (with special focus on change-of-control), supplier and manufacturing contracts, employment and non-compete architecture, IP filings and enforcement history, and litigation exposure. Each requires distinct testing. - Change-of-control provisions are the single legal artifact that most often changes deal structure. A material customer with CoC termination rights forces either a consent process pre-close (delays) or a structural workaround (escrow, indemnity carve-out). Miss the CoC and the workaround stays undesigned. - Employment agreements and non-competes hold retention consequences (see Part IX). Legal DD tests whether the non-competes are enforceable in the target’s jurisdiction (California, Massachusetts, and post-2024 FTC rule effects vary) and whether the target’s equity plan provides for accelerated vesting on transaction. - IP filings audit tests whether patents cover the actual products, whether trade secret protection procedures exist and are being followed, and whether the freedom-to-operate analysis has been run. Patent count alone falls short as the test; enforcement history and coverage against roadmap products matter more. - Three recurring failures: change-of-control provisions catalogued but not tested for consent risk, non-compete enforceability assumed; verification stays the harder discipline, IP filings audited without freedom-to-operate testing. ## The contract is the legal encoding of the commercial thesis The first nine workstreams told the deal team what the target has to be, needs to make, expects to sell and depends on. WS10 is where those commercial, technical and organisational claims meet the contractual paper that governs them. The workstream succeeds when it converts each material claim from the preceding workstreams into a specific contractual position the buyer can point to and rely on; it fails when it accepts management's characterisation of the commercial relationships without testing that characterisation against the executed contract text. **Legal DD runs on risk allocation; risk identification stays the lighter output.** Legal counsel identifies the legal exposure in a contract. The DD workstream converts each identified exposure into a specific mechanism that assigns the exposure to a specific party (seller, buyer, insurer, escrow, financing party, customer or supplier). A legal risk that is identified without being assigned stays uncovered by underwriting; it has been noted. ### The three killer lines the workstream should own - **The CIM tells you what management says the business is. The contract tells you what the business is entitled to.** - **A commercial assumption that cannot survive the contract review is an exposure.** - **Legal DD does not end when a risk is identified. It ends when someone is assigned the risk.** ### Paper versus reality: the recurring test WS10 is where narrative meets paper. Screen every material commercial claim from the preceding workstreams against the executed contract text and identify where the two diverge. The divergence, in either direction, is the DD finding. | Management says | Contract says | DD interpretation | | ------------------------------------------------------ | ------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------ | | Strategic long-term customer partnership | Purchase orders terminable at will on 30 days notice | Revenue is at-will; WS04 retention model overstates durability | | Multi-year committed backlog of $500M | Purchase orders that are non-binding forecasts | Backlog signals intent; commitment stays the harder claim; WS08 must weight it accordingly | | Proprietary technology | Critical software licensed from a former employer with change-of-control termination | IP position is contingent; WS05 architecture assessment must be revisited | | Strategic supplier relationship with reserved capacity | Framework agreement with allocation subject to supplier discretion | Capacity stays unreserved; WS06 capacity entitlement is weaker than represented | | Long-term hyperscaler master agreement | Master agreement with unilateral customer termination on change of control | Deal-critical customer relationship terminates on close absent consent | ### The contracted-revenue trap Aggregated backlog numbers are one of the easiest ways for the commercial narrative to overstate contractual reality. Classify every reported backlog dollar against a five-stage taxonomy and price the underwriting against the composition; aggregate framing understates the risk distribution. 1. **Contracted.** Legally binding customer commitment to purchase named volume at named price within a named period. Enforceable. 2. **Ordered.** Purchase order exists but cancellation rights matter. Value depends on cancellation notice period and any customer-side penalties. 3. **Forecast.** Customer has indicated expected purchase volume without contractual commitment. Weight against customer's own capex track record. 4. **Pipeline.** Sales-team-identified opportunity with no customer-side commitment. Zero contractual weight. 5. **Strategic relationship.** Long-standing counterparty with no enforceable purchase commitment. Zero contractual weight; retention-mechanism analysis (WS04) is the relevant framework. Management's aggregate backlog number is a weighted average across these categories. The DD workstream produces the composition and lets WS08 apply the appropriate confidence-weighted value to each tranche. ## Why hyperscaler paper restructures the legal DD scope The generic legal-DD territory (corporate, contracts, IP, litigation, employment) exists in every industrial-tech transaction. What changes for AI-infrastructure targets is where the risk concentrates. Four hyperscaler-specific features skew the workstream from routine review to finding. First, buyer concentration. When two or three hyperscalers account for 40-70 percent of target revenue, one change-of-control clause in one master supply agreement can gate the entire transaction. That is a different order of consent risk from a diversified industrial customer book where any single consent failure is absorbable. Second, hyperscalers use CoC provisions more assertively than industrial customers. Terminate-on-acquisition, right-of-first-refusal on the equity, mandatory re-price windows, and audit-and-inspection rights that survive change of control all appear in current hyperscaler paper. These are commercial leverage points and the hyperscaler procurement team knows how to use them at consent time; standard boilerplate stays the weaker form. Third, multi-year capacity commitments amplify the exposure. A five-year take-or-pay for capacity or a multi-year GPU allocation contract holds a CoC provision whose economic value sits in the discounted stream; current-quarter revenue stays the lighter measure. Losing a hyperscaler consent means losing the entire remaining commitment; current-period loss stays the lighter case. Fourth, AI-specific IP indemnity language is evolving fast. Foundation-model training-data indemnities, model-output IP-infringement carve-outs, and open-source model licensing terms appear in 2025-2026 hyperscaler paper in forms that did not exist two years ago and that R&W insurers are still learning to underwrite. The DD workstream should read these clauses freshly; template-held assumptions stay unreliable. The practical implication: for AI-infrastructure targets, the material-contracts sub-workstream needs a named senior lawyer with hyperscaler-paper experience assigned in Week 1; associate-template delegation stays weaker. What follows in the rest of this essay is the generic territory; the hyperscaler-specific readings above are on top of it. ## Corporate structure and cap table Corporate DD starts with the entity structure. What entities exist, in what jurisdictions, in what ownership relationship. Simple structures (single holding company with operating subsidiaries) require light review; complex structures (multiple layers, offshore vehicles, prior sponsor structures) require more. Cap table review confirms who owns what, including options, warrants, convertible notes, and any preferred equity with liquidation preferences. For prior sponsor-owned targets, the preferred equity waterfall determines who gets what at exit; the DD test is whether the price at closing meets or exceeds the preferred waterfall threshold. ## Material contracts Material contract review covers customer, supplier, licence, and financing agreements. The DD test for each is: what changes at transaction (change-of-control provisions), what holds forward (assignable terms), and what expires or renews in the coming 12-24 months. Financing agreements deserve a specific flag: existing debt and lease facilities typically hold CoC / assignment provisions that trigger consent, prepayment, or renegotiation at close, and the mechanics of what breaks and what has to be replaced are in Part XIII. The legal workstream surfaces the clauses; the financing workstream sizes the replacement. Customer change-of-control provisions hold the highest impact. A material customer with the contractual right to terminate on acquisition can force pre-close consent processes that delay the transaction by weeks or months. Larger customers sometimes use consent as leverage for pricing or term concessions. The DD test surfaces these early enough for the legal team to negotiate a workaround (escrow, indemnity, price adjustment) or for the deal team to walk. The concentration and power-dynamic reading that determines whether the customer will exercise the consent leverage is in Part IV customer and pricing DD; the legal workstream identifies the clause, but Part IV supplies the commercial context that predicts whether the customer will use it. #### Chart 1\. Change-of-control provisions in material contracts: impact by customer size CoC provisions with material customers are the highest-impact contractual finding in legal DD. Provisions with smaller customers matter less because the customer lacks leverage to enforce. Impact indicative from post-close deal-structure analyses. ## Employment agreements and non-competes Employment DD tests employment agreements, non-competes, and equity plan provisions. For industrial-tech targets, the non-compete architecture matters most because engineering and sales leaders hold direct commercial value. Non-compete enforceability varies by jurisdiction. California prohibits post-employment non-competes almost entirely. Massachusetts requires garden leave payment during the enforcement period. The FTC’s 2024 rule attempting to ban non-competes nationwide was blocked; the FTC voted in September 2025 to drop its appeals, and the rule was formally removed from the CFR in February 2026, reverting federal enforceability to the pre-rule regime. Multi-state operators need jurisdiction-specific review. Equity plan review tests what happens to unvested equity on transaction. Standard PE structures require accelerated vesting on change of control for management team; without acceleration, key people can face material equity loss that undermines retention structure. ## R&W insurance market: what DD findings drive policy structure R&W insurance is standard on mid-market and large-cap industrial-tech deals. Structure, tower design, premium benchmarks, and escrow interaction are in Part XII; here the treatment is narrower: how the DD workstream determines the R&W insurer can underwrite. An R&W insurer will not cover risks that DD identifies as material and unresolved, but the DD process also determines the insurer is willing to underwrite in the first place. The causality runs both ways, which is worth naming. DD findings drive the R&W policy: environmental liabilities that DD Phase I identifies but Phase II has not yet confirmed are in an underwriting grey zone (elevated deductibles, exclusion, or Phase II required before binding); IP infringement claims active at signing get excluded from R&W and require separate escrow carve-out; cyber breach disclosure gaps can void coverage if discovered post-close. But the reverse is also true: the known availability of R&W for certain risk categories influences the scope and intensity of the DD itself. A DD team that treats environmental review as pro-forma because R&W will cover it is running a shallower workstream than one that treats R&W as underwriting the residual risk after DD has done its work. The correct position is that DD sizes the risk; R&W transfers the residual. Detailed premium benchmarks (currently in the 2.5-4 percent range per major broker guidance for standard risk), survival periods (typically 6 years fundamental / 3 years general, with variation), coverage sizing (10-15 percent of enterprise value for uninsured deals; 0.5-1 percent retention + insurance tower for the insured model that now dominates), and multi-carrier layering at scale (see Supplement B on Eaton / Boyd Thermal for the $9.5B example) all are in Part XII. #### Chart 2\. R&W insurance underwriting: how DD findings drive policy structure Environmental findings, IP disputes, and cyber breach patterns all determine R&W insurers will cover, exclude, or require additional evidence for. The DD workstream and the insurance broker should coordinate in Weeks 3-4 of Stage 2. Coverage patterns from Marsh, Aon, WTW published R&W market data 2020-2026. ## IP filings and freedom to operate IP legal DD tests three things beyond the technical IP audit in Part VI. The IP legal audit first examines patent enforcement history, including whether the target’s patents have been asserted and with what outcome. It then reviews trade secret protection procedures to confirm documented processes are being followed; written-down treatment stays weaker evidence. Finally it conducts a freedom-to-operate analysis to identify third-party IP that restricts what the target can build today or on the roadmap. Freedom-to-operate is where legal and technical DD intersect. A target claiming leadership on a second-generation product needs a freedom-to-operate analysis on the product architecture; competitor patents that read on the architecture surface as customer-driven infringement claims post-close. #### Chart 3\. IP legal DD scope: three tests Patent enforcement history reveals whether the portfolio is defensible. Trade secret procedures reveal whether protection exists. Freedom-to-operate identifies what the target cannot build without third-party licence exposure. Effort per test indicative for mid-market industrial-tech target. ## The 12-dimension customer contract audit Customer contract review should test the twelve dimensions; the standard duration + termination + change-of-control triage misses several. Each dimension converts a commercial claim into a specific enforceable (or unenforceable) position. 1. **Revenue enforceability.** Is the revenue contractually committed, or is it forecast, order-based or informal? 2. **Volume commitment.** Minimum committed volume vs maximum optional volume. Two very different numbers frequently reported as one. 3. **Pricing.** Fixed price, indexed price, cost-plus, negotiated per order, or subject to periodic renegotiation? 4. **Exclusivity.** Does the customer have preferential rights? Does the target have exclusive obligations that constrain other customer relationships? 5. **Termination.** For convenience, for cause, for change of control, on notice? Notice period? Wind-down obligations? 6. **Assignment.** Can the contract move to the buyer without customer consent? 7. **Change of control.** Does acquisition trigger consent, notice, termination or economic reset? 8. **Most favoured nation (MFN).** Does any customer receive automatic parity with pricing offered to another? 9. **Liability caps.** What is the maximum contractual exposure? Is it aggregate, per-event or unlimited for specific categories? 10. **Warranty and indemnity.** Who bears failure costs? What triggers customer-side indemnity? What are the target's obligations on IP infringement, product liability, or breach of representation? 11. **IP allocation.** Who owns customer-specific development, jointly developed IP, or improvements to the target's core IP? 12. **Forecast binding.** Are supplied forecasts binding purchase commitments or aspirational planning estimates? ### Take-or-pay depth A take-or-pay clause is materially stronger than "forecasted demand", but not automatically bulletproof. Test each take-or-pay commitment against: - **Duration.** Multi-year with fixed volumes or annual with rollovers? - **Termination rights.** What customer-side conditions release the commitment? - **Force majeure carve-outs.** Broadly defined force majeure can eliminate the take-or-pay obligation entirely under conditions the target cannot influence. - **Capacity reservation.** Is the target obligated to hold specific capacity in return? - **Counterparty creditworthiness.** Enforceable commitment from a financially distressed customer falls short of cash certainty; feeds directly to the customer power framework in WS04. - **Damages.** Liquidated or actual? Capped or uncapped? - **Limitation of liability.** A take-or-pay commitment with a per-event liability cap below the annual commitment value is functionally optional. ### Contractual strength versus economic strength A contract can be legally enforceable and economically weak. The DD test asks not only whether the target has the legal right, but whether exercising that right produces the economic outcome the underwriting assumes. - Damages caps below the true economic loss (contract worth $50M annually with a $2M annual liability cap). - Customer termination rights that are exercisable at commercially reasonable notice. - Broadly defined force majeure that transfers most execution risk to the target. - Enforcement cost that exceeds the recovery, particularly in cross-border contexts. - Supplier dependence on the customer that makes practical enforcement commercially unviable even where legally supported. - Customer set-off rights against unrelated obligations that dilute the target's recovery. ### Remedies asymmetry Ask the same question in both directions. If the target fails to perform, what can the customer recover? If the customer fails to perform, what can the target recover? Frequently, the answers are asymmetric, with the customer holding broad practical remedies (termination, offset, reputational leverage, non-renewal) and the target holding only limited contractual damages. The asymmetry is the risk allocation; the individual damages cap stays the lighter measure. ### Liability caps against physical risk Liability caps only protect against direct contractual damages. The economic consequences that live outside the cap frequently matter more than the cap itself. A $1M contractual liability cap on a $10M annual customer contract may look protective, but a product failure that causes $50M of customer downtime produces termination, non-renewal, indemnity claims, reputational harm and commercial pressure that the cap does not address. Legal DD reads the cap; investment DD reads what remains outside the cap. ## The IP 7-question framework IP DD should decompose into distinct questions; a consolidated "does the target own the IP" assessment misses the structure. Each question produces a different DD finding and a different transaction consequence. 1. **Ownership.** Does the target legally own the IP, or does it license, jointly own, or hold contested rights? 2. **Validity.** Are the patents, trademarks and registrations valid, in force, and free from pending challenges? 3. **Scope.** What do the specific claims protect? A patent count falls short of a scope statement. 4. **Encumbrance.** Is the IP pledged as security, subject to licence-back obligations, or restricted under existing agreements? 5. **Freedom to operate.** Can the target commercialise the product without infringing third-party rights? The DD test requires a specific FTO analysis; a general search misses the enforcement risk. 6. **Transferability.** Can the IP transfer to the buyer, or does it hold consent requirements, change-of-control restrictions, or joint ownership complications? 7. **Employee invention assignment.** Did the specific employees who invented the core IP execute enforceable assignment agreements? Gaps in employee assignment are one of the most common hidden IP issues. ### Employee IP assignment and contractor code The most damaging IP findings frequently stand outside the patent register. A brilliant proprietary technology can be undermined by a key invention developed before employment (pre-existing IP rights the employee retained), contractor-created code without proper assignment (contractor retains ownership by default in some jurisdictions), university-developed IP where sponsored-research clauses give the university ongoing rights, or joint-development-agreement IP where the counterparty retains licence-back or full ownership rights. Each of these can survive an otherwise-clean IP register audit and materially degrade the technology thesis. ### Open-source software (OSS) audit Hardware and industrial-technology targets with embedded software require an OSS audit that most legal DD scopes miss. The specific questions: - What OSS is embedded in the target's product stack? - Under what licences (permissive vs copyleft)? - Are copyleft obligations (GPL, LGPL, AGPL) triggered by the target's distribution model? - Are source-disclosure obligations at risk of being triggered post-close under the acquirer's distribution model? - Are OSS dependencies tracked in a software bill of materials, or opaque to the target's own engineering leadership? A target that describes its software as "proprietary" while embedding substantial copyleft OSS may face source-disclosure obligations that materially compromise the technology thesis. This is where legal DD meets WS05 technology DD directly. ## Supplier contract audit: does the paper reserve the capacity? WS06 established supply-chain dependence. WS10 asks what contractual protection exists around that dependence. For each critical supplier relationship, the specific questions: - Allocation rights (is a specific volume reserved for the target, or is allocation subject to supplier discretion?) - Minimum volumes and take-or-pay symmetry (does the supplier owe volume, or does the target owe volume?) - Lead times contractually committed or supplier discretion? - Price escalation formula (cost pass-through, indexed, fixed, negotiated?) - Capacity reservation payment terms and cancellation rights - Quality standards and rejection rights - Warranty obligations and remediation timelines - Termination rights (supplier-side and target-side) - Change of control (does supplier consent required on acquisition?) - Dual-sourcing rights or restrictions - Force majeure carve-outs ### Does the contract reserve the capacity management says it has? The single most useful supplier-contract test for AI-infrastructure targets. Management represents: - "We have secured transformer capacity for our 2027 build-out." Contract: framework agreement with no binding allocation. - "Our supplier has committed to production volume." Contract: rolling forecast, no purchase commitment on supplier side. - "Long-term semiconductor supply agreement in place." Contract: supplier may reallocate during industry-wide shortage. - "Reserved packaging slots for the ramp period." Contract: reservation payment made, but priority ranking subject to supplier discretion. Each of these distinctions can invalidate the WS08 revenue projection. A supply agreement that guarantees 60,000 units where the projection assumes 100,000 units sits above a supply issue; it is a WS08 modelling error caused by treating a legal claim as a legal reality. ## Change of control as the connective tissue of WS10 Change of control is the pivotal recurring provision because an acquisition changes ownership, governance, financing, control and sometimes jurisdiction, and every material contract can react. The DD workstream classifies every material contract against its CoC disposition. 1. **No impact.** Contract continues on identical terms post-close. 2. **Notice required.** Buyer must notify counterparty; no consent right. 3. **Consent required.** Counterparty must consent; consent may be conditional or unconditional. 4. **Termination right.** Counterparty may terminate (without being required to). 5. **Automatic termination.** Contract terminates by operation of the CoC provision. 6. **Economic reset.** Contract survives but with adjusted pricing, volume or terms triggered by CoC. The output feeds directly into the contractual concentration map below and into WS12 conditions precedent. ## The contractual concentration map The output of WS10 that most directly informs WS12 is a contractual concentration map: for each material contract (top customers, top suppliers, critical licences, financing agreements), the revenue or dependency exposure, the CoC disposition, the termination rights, the consent requirements and the resulting risk classification. | Contract | Revenue / dependency exposure | CoC disposition | Termination | Consent required | Risk classification | | ----------------------------------- | ----------------------------- | ---------------------------- | ------------------------------------------ | ------------------------------ | ------------------- | | Customer A (hyperscaler) | 25% of revenue | Consent required | Customer termination on CoC absent consent | Yes | High | | Customer B (enterprise) | 15% of revenue | Notice | Limited | No | Low | | Supplier C (critical single-source) | Critical Tier-2 dependency | Consent required | Supplier termination right | Yes | High | | Supplier D (multi-sourced) | 10% of BOM | Notice | Limited | No | Low | | IP licence (embedded software) | Product-critical dependency | Automatic termination on CoC | Not applicable | Yes (novation required) | Critical | | Financing agreement | Debt facility | Change-of-control put | Lender right | Yes (refinancing or amendment) | Managed via WS13 | High-risk contracts convert directly into closing conditions in WS12\. Consent-required contracts become the pre-signing legal workstream deliverable. Automatic-termination contracts require novation, replacement or restructuring before signing. ## Exclusivity and MFN as commercial constraints Contract categories that frequently escape standard legal DD and materially affect the WS04 commercial thesis: **Exclusivity.** Contracted exclusivity to a specific customer or supplier can be strategically valuable (locks in a scarce counterparty) or strategically costly (contracts away flexibility to serve competing customers, enter adjacent markets or use alternative suppliers). Screen every exclusivity clause against the WS04 target market and pricing model. **Most favoured nation.** An MFN clause with a large customer means that any price increase offered to a new customer automatically applies to the MFN counterparty. WS04 may conclude the target has strong pricing power; WS10 may reveal that pricing power is bounded by MFN parity obligations. The two must be reconciled or the pricing plan is unenforceable. ## Contractual IP rights: customer-funded development Where the target has developed IP under customer-funded NRE or joint-development agreements, the IP ownership treatment materially affects the "proprietary technology" claim. Common patterns to test: - Customer owns the funded IP outright, with target retaining licence-back rights of varying scope. - Joint ownership with cross-licence. - Target retains ownership but grants customer a perpetual, royalty-free, sub-licensable licence (functionally equivalent to customer ownership for competitive purposes). - Field-of-use restrictions that limit the target's ability to commercialise the IP in the customer's own markets. WS05 architecture DD calls a technology proprietary; WS10 checks whether the contracts preserve that proprietary status. ## Litigation exposure Litigation review covers active litigation, threatened litigation, and settled claims that could recur. For industrial-tech targets, patent infringement and product liability are the categories that hold material exposure. The DD test asks: what are the claims, at what asserted damages, at what estimated exposure, and what R&W insurance treatment applies? Litigation that will be resolved post-close needs indemnity carve-outs; litigation that could produce material damages needs price adjustment. ## Three recurring failures Change-of-control provisions catalogued but not tested for consent risk. The DD team lists which contracts have CoC provisions but does not model the consent process required to close the transaction with the customer relationship intact. Post-close, the customer either withholds consent (forcing termination) or extracts concessions. Non-compete enforceability assumed; verification stays the harder discipline. The DD team confirms non-competes exist but does not test enforceability in the target’s jurisdictions. In California, the non-compete itself is almost certainly unenforceable regardless of what the document says, though customer non-solicits and trade-secret protections remain enforceable and often hold most of the practical retention effect. Retention structure that depends on the non-compete alone has a false floor. IP filings audited without freedom-to-operate testing. The DD lists patents held and confirms status but does not test whether third-party patents restrict the target’s roadmap. Post-close, competitor infringement claim or forced licence surfaces as material cost. #### Chart 4\. Three recurring legal DD failures CoC consent risk not modelled is the most impactful. Non-compete enforceability assumed is the second. FTO not tested is the third but often the highest-cost surprise. Frequency from post-close legal DD retrospectives. ## A worked example Continue the magnetics acquisition. Legal DD scope: Corporate structure: US Delaware holding company, US operating subsidiary, Malaysia operating subsidiary. Clean structure. Cap table: founder 45 percent, employees 22 percent, prior seed investor 8 percent, prior Series A investor 25 percent. Preferred equity waterfall: seed and Series A have 1x non-participating preference; at $400M valuation all classes converted to common. Clean. Material contracts. Two hyperscaler customer contracts: Hyperscaler A has no CoC provision; Hyperscaler B has a CoC notification requirement (no termination right, just notice). Third-party OEM customers: three of twelve have CoC provisions, all standard termination-on-notice. Consent process required for two Hyperscaler B contracts and three OEM contracts; manageable within 8-10 week Stage 4 window. Employment. Founder non-compete: 4 years, national scope, Malaysia enforceable but California unenforceable (founder resides in California). Structure risk on Malaysia operations. Engineering leads: 12-month non-competes in Malaysia; enforceable. US engineers: California unenforceable. Retention structure adjusted accordingly (rely on rollover equity and retention bonus for CA-based engineers, non-compete for Malaysia-based). IP. 47 patents. Enforcement history: two prior assertion attempts (both settled). Trade secret procedures documented for ferrite formulation. FTO analysis identifies one competitor patent that reads on second-generation product; licence available at estimated $2-4M annual royalty. Deal structure needs to accommodate. Litigation. One active patent infringement claim (target defending), asserted damages $8M, R&W insurance treatment available. One product liability claim in Malaysia (settled for $1.2M). No material threatened litigation. #### Chart 5\. Worked example legal DD scorecard Corporate + cap table clean. Contracts require consent process on 5 items. Employment has jurisdiction-driven risk. IP requires FTO licence structure. Litigation manageable. Composite scorecard for the worked example. ## Litigation as strategic threat; dollar-list framing misses the strategic consequence The conventional DD output ("12 open litigation matters") falls short of useful for underwriting. The investment question is what each matter threatens; the aggregate count misses the per-matter severity. Screen each material matter against the threat categories: - **Cash.** Direct damages exposure and defence cost. - **IP.** Loss of specific IP rights or introduction of licence-back obligations. - **Customer relationship.** Damages to specific customer relationships as a consequence of the litigation itself. - **Regulatory licence.** Loss of a specific licence, authorisation or approval required to operate. - **Manufacturing continuity.** Injunctions or restrictions on specific manufacturing activity. - **Reputation.** Damage to the target's market credibility affecting future customer decisions. Materiality spans beyond financial dimensions. A $2M lawsuit threatening core IP is materially more consequential than a $20M routine commercial dispute. The workstream classifies each matter by probability, exposure and strategic consequence; aggregate dollar value alone misses the structure. ## The WS10 output: 9 discrete deliverables 1. **Contract universe.** Every material agreement mapped and classified. 2. **Economic rights.** What each party is contractually owed and by when. 3. **Control rights.** What happens on change of control, with per-contract CoC disposition. 4. **IP rights.** Per-IP-asset ownership, validity, scope, encumbrance, transferability status. 5. **Liability allocation.** Who bears loss for each identified risk category. 6. **Termination and concentration.** Which contractual relationships can disappear under what triggers. 7. **Litigation and disputes.** Per-matter classification against the six-threat framework. 8. **Contractual gaps.** Where the paper fails to support a specific commercial assumption from WS03-WS09. 9. **Transaction consequences.** Each finding translated into the specific WS12 or WS13 mechanic that assigns the risk. ## WS10 hands off to WS12 and WS13 | WS10 finding | Downstream mechanism | Buyer protection | | -------------------------------------------------------------- | --------------------------------------------- | -------------------------------------------------------------------------- | | Customer CoC consent required on top-3 customer | WS12 closing condition | Consent as condition precedent to close | | Supplier CoC termination right on critical single-source input | WS12 closing condition + WS13 MAC clause | Consent or replacement supplier as CP; MAC protection if consent fails | | IP licence with automatic CoC termination | WS12 novation or replacement | Novation pre-signing; specific IP indemnity | | Employee IP assignment gap on core invention | WS12 specific indemnity + retention | Uncapped IP indemnity; retention of named inventor | | Contractual liability cap materially below economic exposure | WS12 R&W insurance + WS13 minimum liquidity | R&W policy coverage; liquidity buffer for uncovered downside | | Take-or-pay commitment with broad force majeure carve-out | WS08 revenue-plan adjustment + WS12 CP | Reduced revenue-plan weight; potentially closing condition on FM amendment | | MFN clause bounding pricing power | WS04 pricing-plan revision + WS08 stress case | Revised pricing plan; specific downside case in WS08 | | Litigation threatening core IP or regulatory licence | WS12 escrow + specific indemnity | Escrow proportional to potential damages; indemnity for specific matter | | Corporate-structure or cap-table cleanup required | WS12 pre-signing legal workstream | Pre-signing restructuring; specific R&W on capital structure | | Financing agreement CoC put | WS13 refinancing structure | Refinancing planned pre-close or committed backup facility | Every row is a specific WS10 finding translated into a specific WS12 or WS13 mechanism. Legal DD that stops at "here are the legal risks" leaves the underwriting undefended at the layer where the risks are supposed to be allocated. **The WS10 contrarian.** The contract is the legal encoding of the commercial thesis. When contract and thesis disagree, the contract governs the economics; the thesis governs the pitch. Legal DD runs on risk allocation; risk identification stays the lighter output: the workstream succeeds when every material commercial claim from WS03-WS09 has been tested against the contract text and every identified risk has been assigned to a specific party through a specific mechanism in WS12 or WS13\. That is transaction architecture. It is materially different from the "legal DD is a review of material contracts" playbook that generalist PE DD firms most commonly deliver. ## Closing Legal DD is executed almost entirely external but requires internal counsel to own the interpretation. The five deliverables that need to land: contract review + CoC risk map, employment + non-compete jurisdiction analysis, IP + FTO analysis, litigation exposure with R&W treatment, corporate structure clean-up requirements. #### Chart 6\. Legal DD checklist Five deliverables by end of Stage 2. 01 Material contract review + CoC map Customer, supplier, licence, financing 02 Employment + non-compete review Jurisdiction-specific enforceability 03 IP + FTO analysis Patents, trade secrets, third-party restrictions 04 Litigation exposure + R&W Active, threatened, indemnity treatment 05 Corporate structure + cap table Entity map, waterfall analysis Legal firm (Kirkland, Latham, Skadden or equivalent) executes all five with in-house counsel oversight. Part XI moves into regulatory, environmental, cyber and ESG DD. ### Glossary of terms used CFIUS Committee on Foreign Investment in the United States. US inter-agency review body that screens foreign acquisitions of US businesses for national security implications. CoC Change of Control. Contractual clauses that trigger renegotiation, consent requirement, or termination when a counterparty is acquired or its ownership changes. DD Due Diligence. The workstream discipline of testing a target's claims before committing to a transaction. HSR Hart-Scott-Rodino Antitrust Improvements Act. US merger notification threshold and review regime. R&W Representations and Warranties. Seller statements about the target's condition. R&W insurance covers buyer's post-close claims arising from breaches. For the full corpus glossary of acronyms used across all essays, see [adikumar.co/glossary](https://adikumar.co/glossary/). Related topic hubs - [AI Infrastructure M&A](https://adikumar.co/ai-infrastructure-m-a/) - [AI Infrastructure Financing](https://adikumar.co/ai-infrastructure-financing/) - [Browse all topic hubs →](https://adikumar.co/topics/) ### Sources and method FTC non-compete rule status per 2024-2026 litigation. Patent enforcement history per USPTO records. R&W insurance treatment per major broker (Marsh, Aon, WTW) published guidance. ### Series Due Diligence for the AI Buildout. Part X (this piece): *Legal and contract DD*Written in a personal capacity. Nothing here is investment advice. ## Contract value-at-risk matrix The synthesis output of contract DD. Every material contract gets scored on the dimensions that determine whether a change of control triggers value loss. The matrix template below holds into the transaction structure workstream, where each unresolved cell becomes a rep, a warranty, an escrow, a condition precedent, or a walk. Populate against the target’s actual contract set during DD; the columns are constant, the rows are target-specific. | Contract | Revenue exposure | Change-of-control | Termination | Assignment | Security / collateral | Value at risk | | -------------------------------- | ----------------------------- | ---------------------------- | ------------------------------- | ----------------------------- | --------------------------- | --------------------------------------- | | Hyperscaler capacity lease | % of ARR | Consent / termination on CoC | Notice period, penalty basis | Assignable to affiliates only | Standby LC, cash deposit | $ if CoC triggers renegotiation or exit | | PPA (long-tenor) | Power cost hedge; margin lock | Utility consent typical | Force majeure, curtailment | Regulatory constraint | Parent guarantee | $ if PPA re-priced or terminated | | Grid interconnection agreement | Enables all revenue | ISO/RTO consent required | Firm vs non-firm exposure | Queue-position transfer rules | Milestone deposits | $ if project fails to energise | | Tax abatement / incentive | After-tax IRR uplift | Local authority consent | Clawback on default | Personal to grantee | Performance bond | $ present value of lost abatement | | Key vendor / supply agreement | Cost lock; supply security | Consent, MFN reset | Volume commitments, take-or-pay | Assignment permitted | Deposits, parent guarantees | $ margin loss on renegotiation | | Debt facility / credit agreement | Capital structure | Mandatory prepayment on CoC | Event of default triggers | Lender consent required | All-asset security package | $ prepayment cost + refi risk | Each row that reads “consent required” on change-of-control is a Week-1 diligence request: the buyer needs to know before signing whether consent is likely, whether the counterparty is likely to reprice on consent, and whether the deal can survive without that specific contract. Every “value at risk” cell that lands above the deal’s materiality threshold becomes a specific reps-and-warranties construct, an escrow slot, or a closing condition in Workstream 12. Due Diligence for the AI Buildout. The workstream index 1. [Part I. Framework: how DD works for AI infrastructure deals](https://adikumar.co/due-diligence-for-the-ai-buildout/) 2. [Product, technology, and roadmap DD](https://adikumar.co/product-architecture-and-technology-dd/) 3. [Customer, pricing, and competitive DD](https://adikumar.co/customer-pricing-and-competitive-dd/)