Private Equity AI Report, November 2026

The decision in brief

Underwrite AI benefits only to the extent that their economics, operating dependencies and rights survive the transaction. Separate benefits already in the earnings base from incremental improvements, then deduct the recurring resources needed to sustain them and identify transition cash costs. The acquisition case should make clear what the buyer receives, what must be rebuilt and who can authorize the required changes.

Executive summary

An AI initiative can be useful to the seller and still require substantial work after acquisition. The workflow may depend on a parent-company agreement, a key employee, a shared data environment or a review process that is outside the deal perimeter. A diligence paper that considers only the software and the claimed time saving can miss those dependencies. The buyer needs to understand whether the benefit is transferable and what maintaining it will cost.

OFF's November evidence supplies relevant questions, not a private-equity performance benchmark. In the updated investor sample, 141 of 364 respondents selected cutting costs as an area of portfolio impact. In the carried-forward operator sample, 34 of 77 reported real-time visibility into AI running costs, 30 partly and 13 no. These separate self-reports do not establish the prevalence of cost savings in PE-owned companies or the completeness of any target's financial controls. [INVIMPACT; COST]

The report's central distinction is between the seller's reported benefit and the buyer's incremental, sustainable result. A benefit may already be reflected in the earnings base, may represent unused capacity rather than a cash reduction, or may require continuing costs omitted from the seller's presentation. A transaction can also change the cost or availability of the resources on which the benefit depends. Those questions belong in the operating and financial case before they become post-close surprises.

November contributes a benefit-and-remediation bridge, a transferability matrix and an illustrative ownership-transition review. The bridge keeps recurring economics separate from one-time transition spending. The matrix identifies which rights, services, people and records must remain available. The review distinguishes a control acquisition from a minority investment, where the investor may have limited authority to implement its preferred operating plan.

A detailed AI diligence process can delay a transaction without resolving the uncertainty that affects its value. The recommended approach is to focus on the few AI claims that materially affect the investment case or continuity of operations. Each diligence request should identify the decision it can change. The purpose is to qualify an acquisition assumption and fund the work required to sustain it, not to demand a comprehensive technical audit of every tool in the target.

What November adds

Earlier OFF dealmaker material established that AI claims require evidence beyond adoption. This edition develops the ownership-transfer problem in more detail. A workflow's value before closing does not establish its value to the buyer after closing, especially when the transaction changes contractual access, management authority or the operating environment. The new analysis follows the benefit through that change.

The first addition is a disciplined incremental bridge. It distinguishes a claimed gross benefit, capacity without a demonstrated cash consequence, benefits already included in the earnings base, recurring costs and new transition spending. These items should not be blended into one adjustment. The bridge is a diligence tool; it does not prescribe accounting treatment or determine which adjustments a buyer, lender or auditor will accept.

The second addition is a transferability assessment with explicit dependencies. It asks whether the buyer can continue using the relevant data, services, integrations and operating knowledge under the proposed transaction structure. A contract, an architecture diagram and a management assurance may each answer only part of that question. The operating team should identify what will actually be available on day one and what requires a separate agreement or replacement.

The third addition is an evidence record that can survive integration and later exit. A buyer should preserve the assumptions used at entry, the changes made after closing and the basis for any subsequent benefit claim. Otherwise, a future seller may present a result without being able to distinguish acquired performance from improvements created during ownership. This is a provenance problem as much as a reporting problem.

The report avoids reusing the October newsletter's financing and multiple-compression example. Its worked case uses a different question: how much of a claimed annual AI benefit remains incremental after qualification and ownership-transition costs? No deal valuation, multiple effect or observed return is inferred from the OFF survey.

Evidence and sample boundaries

The available OFF instruments cover executive buying, security, marketing, investor impressions and a smaller operator sample. They do not provide a dedicated, validated PE sample linked to portfolio-company financial statements. This report therefore uses the measures to identify diligence questions and labels the transaction analysis as synthesis.

The investor impact question has 364 responses and permits multiple selections. Its unit is the respondent's view across a portfolio, not a verified company outcome. The operator cost question has 77 responses and asks about real-time visibility into what AI costs to run. It does not establish workflow-level attribution, completeness of labor costs or compliance with an accounting policy.

The maturity instrument has 91 responses, of which 12 selected embedded, defined in the questionnaire by effects on cost structure or hiring plans if AI were removed. That wording makes operating dependence relevant to diligence, but it does not measure how often dependencies fail or whether a particular target can sustain its result after acquisition. [MATURITY]

All measures are selective, self-reported and cross-sectional. The report does not combine separate populations into a control-gap estimate. External sources provide accounting, cost-management, governance and finance context within their stated scope. The financial bridge below is hypothetical and excludes taxes, financing, working-capital changes and transaction-specific accounting conclusions.

Exhibit 1. Evidence translated into diligence questions

Exhibit note: Separate instruments and populations. Investor impact is multiple selection; cost and maturity are single selection. These results do not estimate PE portfolio-company performance or prove a target-specific weakness.

1. Establish the transaction perimeter of the workflow

Begin with the business service that produces the claimed benefit. Identify the inputs, decisions, actions and people required to deliver it. Then compare that operating system with the transaction perimeter. The buyer may acquire the relevant employees and software contract while excluding a parent data source or shared support team. The diligence question is whether the resulting business can still perform the service under the proposed structure.

A perimeter review should cover rights as well as technical access. The fact that a target can currently retrieve information does not establish the buyer's right to continue every use after closing. Contractual restrictions, customer commitments and the proposed ownership structure require review by the appropriate advisers. The operating team should explain the use in concrete terms so that legal review addresses the actual workflow rather than an abstract reference to AI.

Shared services deserve particular attention in carve-outs. A model gateway, identity service, evaluation process or specialist review team may be supplied centrally. Replacing it can require expenditure, time and capabilities not visible in the target's standalone financial presentation. A transition-services agreement may preserve access temporarily, but the buyer still needs a plan for the end of that arrangement and a realistic view of the replacement cost.

People can be part of the critical infrastructure. A workflow may rely on an employee who knows which outputs require correction or how to resolve unusual cases. That knowledge may not be documented, and the employee may not remain after the transaction. The diligence team should distinguish a repeatable process from a process stabilized by exceptional individual effort. The response may be retention, documentation, training or a narrower initial operating scope.

The buyer should also identify changes it plans to make. Integrating the target into a new identity environment or replacing a supplier can alter the workflow before the expected benefit is realized. Those changes should be considered in the investment case, rather than assumed to be neutral administrative tasks. A benefit supported under the seller's configuration may require renewed evidence after integration.

The output of this chapter's work should be a concise perimeter statement: what transfers, what remains available temporarily, what must be replaced and which dependency is unresolved. Each unresolved item should have an owner and a consequence for the deal or operating plan. A list of technologies is insufficient unless it explains the business result that depends on them.

2. Qualify the benefit before treating it as incremental

A seller may describe several different results as savings. Some may be reduced spending, some avoided hiring, some additional capacity and some an estimate of employee time. The buyer should identify the category before evaluating the amount. A valid capacity improvement can be valuable without constituting an incremental reduction in cash expenditure or an accepted earnings adjustment.

The baseline should be reconstructed from the relevant operating period. If the benefit has already occurred, it may already be included in reported earnings. Adding it again as a prospective improvement would double count the same result. If the benefit is forecast, the buyer should identify the action required to realize it and whether that action is within management's authority and operating plan.

Avoided hiring requires a counterfactual that deserves scrutiny. The target may have planned to hire in response to expected demand, but the demand may have changed. The buyer should understand the workload assumption and whether the hiring would genuinely have occurred without the AI initiative. A cancelled vacancy can support a different claim from a theoretical estimate based on hours saved.

Service quality and workload should be considered alongside cost. A lower expense achieved by handling fewer requests or accepting more rework is different from a sustainable productivity improvement. The benefit record should specify the eligible workload, accepted outcome and relevant quality measure. Without that context, the buyer may mistake a change in service scope for an efficiency gain.

The recurring cost boundary must include the resources needed to sustain the result. Model and infrastructure charges are only part of the question. Review labor, support, data preparation, monitoring and integration maintenance can also matter. The allocation method should be clear enough that the buyer can see which costs are shared and which will change under its ownership.

Qualification should result in a range of supported conclusions, not a forced yes-or-no label. Some benefits may be verified and already in the base; some may be incremental but contingent; some may remain unproven. The investment case should preserve those distinctions. A buyer can choose to fund an uncertain improvement, but it should not describe the expected result as established merely because the seller's presentation uses a precise number.

3. Keep recurring economics and transition cash separate

Recurring operating economics describe the resources required to deliver the service over time. Transition spending describes the work needed to establish or change that operating state. Mixing the two can obscure both the sustainable benefit and the cash required to obtain it. The deal team should show each clearly and explain when the benefit is expected to begin.

Examples of transition work include replacing a shared service, rebuilding an integration, documenting a process or establishing an evaluation record. Some expenditure may create reusable capability; some may simply make the acquired workflow usable by the buyer. The classification and accounting treatment depend on the facts and applicable policies. The diligence model should first identify the work and cash timing rather than assume a favorable treatment.

A run-rate benefit is not the same as first-year cash contribution. The buyer may incur transition costs immediately while benefits begin later or ramp gradually. A model that presents a full annual benefit alongside a short implementation period can overstate the near-term result. The operating plan should identify the prerequisites, sequence and capacity required to reach the stated run rate.

Remediation also competes for management attention. A target may be able to fix several issues individually but lack the staff to do all of them during integration. The deal team should identify the constraint that governs timing. That may be a specialist team, customer approval, data access or a supplier dependency. A realistic plan should not schedule every activity as though those constraints were independent.

The buyer should avoid double counting remediation costs across workstreams. A shared identity project might support both the AI workflow and a wider security integration. The model should identify the allocation and preserve a reconciliation to the total integration budget. Conversely, calling a cost part of the broader integration does not make it disappear from the economic assessment of the transaction.

The resulting bridge should show the recurring benefit after sustaining costs, the transition cash requirement and the timing assumptions. It can then be considered alongside the wider investment case. This report does not prescribe how a lender, auditor or buyer should treat an adjustment. It recommends an evidence record that makes those judgments possible without confusing an operating forecast with an established financial result.

4. Match the improvement plan to actual ownership rights

A control investor can often direct management priorities within the relevant governance and contractual arrangements. A minority investor may have influence and information rights without the authority to implement the same plan. The investment case should reflect that distinction. An attractive improvement that requires decisions the investor cannot cause is a contingent opportunity, not an automatic post-investment benefit.

The improvement plan should identify the required decision-maker for each material action. Examples include changing a supplier, altering a customer workflow, approving a data use or reallocating staff. The investor should understand whether management already supports the action and whether other approvals are needed. A general statement that the investor will bring operating expertise does not resolve those dependencies.

Minority investments may require a narrower evidence approach. The investor can seek agreed reporting and a defined process for discussing unresolved assumptions, while recognizing the company's management responsibilities. The plan should not rely on access to sensitive customer or employee information that the company cannot appropriately provide. Aggregated evidence or a scoped independent review may be more suitable than unrestricted access.

Add-on acquisitions create another transferability problem. A workflow that works in one portfolio company may depend on a particular data structure, customer mix or operating team. Applying it to an acquired business is a new hypothesis. The sponsor should identify what can be reused and what must be adapted before assigning the same benefit rate across the combined group.

Centralized purchasing can improve commercial terms while creating concentration or compatibility issues. The buyer should evaluate both. A portfolio-wide agreement is not automatically the best operating choice for every company, and a local exception may be justified by a specific requirement. The relevant decision should consider total delivery economics, operating risk and the rights needed for the intended use.

Ownership rights should therefore be visible in the benefit case. The model should distinguish improvements management has authorized, improvements requiring further agreement and options the investor may explore. That clarity is especially useful when the investment committee compares opportunities with different control structures. It prevents a persuasive operating plan from receiving more certainty than the investor's actual authority supports.

5. Preserve evidence through integration

Integration can destroy the comparability of the original benefit measure. Systems change, teams merge, workloads shift and cost allocations are revised. Those changes may be sensible, but they can make it difficult to determine whether the acquired workflow delivered the expected result. The buyer should preserve the entry definition and document the changes needed to interpret later performance.

The evidence record should connect the initial claim to the relevant source, operating boundary and financial category. It should identify whether the result was observed, estimated or forecast at entry. After closing, management should record which assumptions changed and why. This allows the team to distinguish a failure of the original thesis from a deliberate change in the operating plan.

Benefits should be measured against an appropriate post-close comparison. If volume increases, a higher total cost may coexist with lower cost per accepted outcome. If the service expands to harder cases, performance may change for reasons unrelated to the original workflow. The operating partner should require a consistent explanation of scope and case mix rather than insist that every headline metric improve monotonically.

The record should also preserve costs that become visible later. A support burden may emerge only after the founder or seller's specialist team leaves. A supplier concession may expire. A manual review process may prove more extensive than expected. These findings should update the economic case and the remediation plan, rather than be treated as unrelated integration noise.

Evidence access should be designed for continuity without unnecessary duplication. The buyer needs enough information to validate the benefit and operate the service, but it may not need unrestricted copies of sensitive source data. The appropriate records, access controls and retention arrangements should be agreed with the relevant functions. A diligence archive should not become an uncontrolled operational data store.

At the end of the initial integration period, the team should reconcile the entry thesis with the operating state achieved. Which benefits are sustained, which remain contingent and which have been withdrawn? What recurring cost is now required? The answer should inform the next allocation of capital and management attention. It should also become the starting point for future reporting, rather than leaving the original investment paper as the only account of what was expected.

6. Build an exit claim from the ownership record

A credible exit narrative should distinguish the performance acquired from the improvements created during ownership. If a benefit was already in the earnings base at entry, it should not later be presented as wholly new value creation. If the sponsor invested in a more reliable or scalable operating system, the evidence should show the change and its economic consequence.

The exit record should connect each material claim to a consistent definition and period. A buyer will want to understand the workload, accepted outcomes, recurring costs and remaining dependencies. An attractive percentage without that context can be difficult to verify. The seller should be able to explain how the measure changed when the business expanded or systems were integrated.

Transferability remains relevant at exit. The next owner may have a different technology environment, supplier agreement or operating model. The seller should identify the rights and services required to continue the benefit, including any arrangements that may need consent or replacement. A benefit that depends on the current sponsor's shared resources should be described accordingly.

The company should also distinguish historical results from the next improvement opportunity. A mature workflow may have delivered a measured cost reduction, while a proposed expansion remains untested. Both can be discussed in a transaction, but they deserve different evidence and certainty. Combining them into one total can make the established result less credible and obscure the work still required.

An unresolved limitation should be disclosed in the appropriate diligence process with enough detail to assess its consequence. The purpose is not to publish sensitive operational information broadly. It is to avoid a situation in which the buyer discovers that a central benefit claim depends on an omitted cost or unavailable resource. A clear record can make the discussion more efficient because it identifies the actual issue to resolve.

This report does not estimate that better AI evidence increases a valuation multiple or guarantees a successful exit. Its recommendation is narrower: preserve the chain from claim to operating result so that a future buyer can assess what it is acquiring. That discipline also helps the current owner make better decisions before exit, because the same record exposes assumptions that no longer hold.

External evidence and its limits

FinOps Foundation guidance on unit economics supports connecting technical costs to a meaningful business unit. For transaction analysis, that suggests examining the cost of an accepted outcome rather than relying only on a model invoice. The benefit bridge below is an original application and is not an industry benchmark or an endorsed adjustment method. [E-FINOPS]

Deloitte's third-quarter 2026 CFO Signals article examines AI cost governance among 200 North American CFOs at companies with at least $1 billion in revenue. Its discussion of visibility, controls and response capability is relevant to operating diligence. The population differs from OFF's selective samples and from many acquisition targets, so its results should not be used to infer a target's control quality. [E-DELOITTE3]

NIST's voluntary AI Risk Management Framework provides a structure for assigning responsibilities and examining risk through a system's context. It can help a deal team organize questions about a material workflow. It does not certify that a target is compliant, that its controls are effective or that a benefit will survive acquisition. Those claims require evidence specific to the transaction. [E-NIST]

The IFRS Foundation's IFRS 15 overview distinguishes customer contracts, performance obligations and the recognition of revenue as obligations are satisfied. Its relevance is to diligence where an AI benefit claim includes new customer revenue. It does not prescribe EBITDA adjustments or apply automatically to every reporting jurisdiction. Accounting advisers must assess the relevant framework and contract facts. [E-IFRS15]

These sources address cost measurement, management practice, risk process and one accounting framework. None supplies a universal AI diligence checklist. The transaction team should use them to sharpen the claims that matter to the deal and identify the evidence required. A broad reference to a recognized framework is less useful than a clear explanation of the unresolved assumption that could change the investment case.

Transactions require decisions before every dependency is resolved. The practical response is to distinguish a condition to closing, a priced or funded transition requirement and a post-close option. The team should know which category each uncertainty belongs in and who accepts the remaining risk. The decision can then be made with a transparent view of the benefit and the work required to sustain it.

A deal team should preserve the provenance of each assumption in its investment paper. If the target provides an observed cost reduction, identify the period and workload. If the buyer proposes a new operating change, identify it as an underwriting assumption with an implementation owner. If an external framework supplies a question, retain it as a diligence prompt until target-specific evidence answers it. This separation matters after signing: the operating team can see which benefits were already demonstrated and which still require work. It also avoids using a well-known research publisher to imply assurance over a transaction the publisher has never examined.

Evidence comparison

An illustrative benefit-and-remediation bridge

Assume a hypothetical target presents $1.8 million of annual AI-related benefit. Diligence identifies $750,000 as staff capacity for which no expenditure reduction or incremental revenue has yet been demonstrated. Another $300,000 is already reflected in the historical earnings base. Removing those items from the incremental claim leaves $750,000 of potential additional annual benefit. This does not reduce historical earnings by $300,000; it prevents counting that existing benefit again.

Assume the remaining benefit is supported by a defined operating plan, but sustaining it requires $250,000 of recurring AI service, review and support cost not included in the original claim. The incremental annual result becomes $500,000. The buyer then identifies $120,000 of additional recurring cost to replace a seller-provided shared service. The modeled incremental annual result under buyer ownership is therefore $380,000 before the other exclusions stated below.

Transition work requires a separate $600,000 cash outlay. If the full annual benefit were available for the entire first year, the simplified first-year cash contribution would be negative $220,000. If the benefit begins only halfway through the year, with a full run rate thereafter, the simplified contribution is $190,000 less $600,000, or negative $410,000. These are timing illustrations, not forecasts or accounting conclusions.

Exhibit 2. Illustrative incremental benefit bridge

Exhibit note: Entirely hypothetical. No EBITDA adjustment, accounting classification, valuation or deal return is asserted. Working capital, taxes, financing, transaction costs and other integration effects are excluded.

The bridge changes the discussion. The target may still be attractive, and the excluded capacity may support future growth. But the investment team should not treat the entire $1.8 million as an immediate incremental earnings improvement. The operating plan should identify how the remaining benefit will be delivered, when it begins and which transition activities are prerequisites.

Exhibit 3. Transferability and ownership review

Exhibit note: Proposed diligence tool. Legal and contractual conclusions require review of the actual transaction documents and circumstances.

The deal team should assign each unresolved item to the person who can resolve it. A legal restriction belongs with counsel and the business owner who understands the intended use. A service replacement belongs with the operating and technology teams. A disputed benefit category belongs with finance and the accountable process owner. The investment committee should receive the combined effect on the case, not separate workstreams that assume each other will solve the problem.

After closing, the same bridge should be updated with observed results. If the benefit starts later, if review cost remains higher or if the capacity supports growth rather than spending reduction, the record should show that outcome. The purpose is to maintain a useful economic account of the investment, not to defend the original number regardless of what the business learns.

The next transaction review

Identify the AI claim that most affects the current deal's operating case. Trace it to a specific workflow, a supported benefit category and the resources required to sustain it. Then compare those resources with the transaction perimeter. This targeted review will often reveal more than a broad inventory of AI tools because it concentrates on the assumption that can change the investment decision.

Keep three questions separate: what is already in the earnings base, what additional result is supported and what cash must be spent to obtain or preserve it? The answer should remain clear when the ownership structure, supplier terms or integration plan changes. A benefit that survives those questions deserves a different level of confidence from an attractive but untested improvement opportunity.

November's recommendation is to make transferability part of the economics from the beginning. The buyer should know what it is acquiring, what it must rebuild and which decisions it can actually cause. That creates a more credible operating plan after closing and a more defensible account of the business when the next owner asks the same questions.

Transaction questions

Is this a PE survey?

No. It is a transaction synthesis using cross-functional OFF evidence. No dedicated PE performance estimate is claimed.

Does reported cost impact prove a realized saving?

No. INVIMPACT records investor impressions, not audited company financial outcomes.

Does running-cost visibility establish complete workflow economics?

No. COST asks about real-time visibility into AI running costs; it does not verify complete allocation or benefit conversion.

Should a benefit already in the earnings base be removed from historical earnings?

Not on that basis. The illustration excludes it only from an additional prospective benefit claim to avoid double counting.

Is the bridge an accepted EBITDA adjustment schedule?

No. It is a hypothetical diligence calculation. Transaction and accounting judgments require the actual facts and relevant advisers.

Does the report estimate an AI valuation premium?

No. It makes no claim about multiples, valuation effects or realized transaction returns.

Citable facts

- Cutting costs selected as portfolio impact: 141/364, 39%, updated investor sample. [INVIMPACT]

- Real-time AI running-cost visibility: yes 34, partly 30, no 13; base 77, carried forward. [COST]

- Embedded maturity: 12/91, 13%, carried-forward operator sample. [MATURITY]

Methods and limitations

Calculation-v1.1 uses authorized 1 November exports through the stated cutoff. Byte-identical exports and invited-only records are excluded. The latest eligible nonblank answer is selected per normalized respondent identifier per question. All distributions retain their own bases and response categories.

The data is selective and self-reported. Separate samples are not linked to transaction outcomes, audited financial statements or identified PE portfolio companies. No causal result, valuation premium or representative PE estimate is asserted. The benefit bridge and transferability matrix are original illustrative tools. Accounting and legal references retain their stated scope.

References

[OFF-DATA] Open Future Forum. Executive AI Statistics, November 2026. INVIMPACT, COST and MATURITY.

[E-FINOPS] FinOps Foundation. Unit Economics capability. https://www.finops.org/framework/capabilities/unit-economics/

[E-DELOITTE3] Deloitte. CFO Signals, third quarter 2026. 21 October 2026. https://www.deloitte.com/us/en/insights/topics/business-strategy-growth/3q-2026-cfo-signals-survey.html

[E-NIST] NIST. AI Risk Management Framework 1.0, current official resource page. https://airc.nist.gov/airmf-resources/airmf/

[E-IFRS15] IFRS Foundation. IFRS 15 Revenue from Contracts with Customers, official overview consulted 2 November 2026. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-15-revenue-from-contracts-with-customers/

Related research and citation

The November CFO report develops benefit qualification, Investor addresses post-investment evidence completion, Investment Banking examines revenue quality, and General Counsel considers authorization and contractual allocation.

Suggested citation: Newlands, Murray. Private Equity AI Report, November 2026, Transaction and Operating Synthesis. Open Future Forum. Evidence cutoff 31 October 2026.

Open Future Forum research: https://openfutureforum.com/research/

Editable working tool

Ownership-transfer benefit bridge: download the editable CSV worksheet. The worksheet implements the report's recommendation; it is not a validated score, mandatory standard or completed assessment.

November report collection

• Executive AI Leverage Report
• CFO AI Leverage Report
• CEO AI Leverage Report
• CISO AI Leverage Report
• CMO AI Leverage Report
• AI Transformation Report
• AI Leaders AI Leverage Report
• Board Director AI Governance Report
• YC Founder AI Report
• VC and CVC AI Investment Report
• Investor AI Report
• Investment Banking AI Report
• General Counsel AI Report
• Executive AI Statistics

Share LinkedIn Twitter / X