Key Takeaways
- U.S. public companies recorded $96 billion in goodwill impairments in 2024, up 16% from $83 billion in 2023, across 273 impairment events among 8,134 monitored companies (Kroll 2025 U.S. Goodwill Impairment Study)
- AI reduces DCF model build time from 4-8 hours to under 10 minutes per reporting unit, compressing the most labor-intensive phase of annual impairment testing (CORAA.ai; multiple vendor data)
- 87% of CFOs at $1B+ companies expect AI to be extremely or very important to finance operations in 2026, yet only 11% have AI fully operational in their finance functions today (Deloitte Q4 2025 CFO Signals; L.E.K. 2025 Office of the CFO Survey)
- A major goodwill impairment audit consumes 60-100 hours of valuation team time under manual processes; AI automation targets the mechanical stages - data normalization, DCF build, benchmarking, and sensitivity runs (CORAA.ai 2025)
- External goodwill impairment study fees run $7,500-$25,000+ per reporting unit engagement; AI pre-processing and automated DCF generation reduce external fees by an estimated 30-50% (Eton Valuation; Sofer Advisors 2025)
AI goodwill impairment automation statistics 2026: what the data shows
Goodwill impairment testing is among the most judgment-intensive tasks in the finance close. Under ASC 350 (U.S. GAAP) and IAS 36 (IFRS), companies must test goodwill at least annually - building discounted cash flow models, estimating WACC, projecting revenue, and allocating goodwill across reporting units or cash-generating units. When a triggering event occurs, that full cycle happens again mid-year.
The mechanical stages of that process - data gathering, DCF model construction, sensitivity runs, benchmarking, documentation - are exactly where AI is starting to reduce workload. The judgment stages - estimating useful lives, setting discount rate assumptions, determining whether a triggering event has occurred - remain human-intensive.
This article covers what the 2026 data shows on AI adoption for goodwill impairment workflows, the cost and time benchmarks involved, the audit implications, and how valuation teams are changing the way they divide work.
For related context, see our AI depreciation automation statistics 2026, AI amortization automation statistics 2026, and AI in accounting and finance statistics 2026.
1. The scale of goodwill impairment: what's at stake
Before examining AI automation, it helps to understand the volume of goodwill that finance teams are managing. The numbers are large, and they explain why companies spend materially on impairment testing.
S&P 500 companies carry roughly $4 trillion in goodwill on their collective balance sheets, near a five-year high (Calcbench; PRWeb 2025). Goodwill as a percentage of total S&P 500 assets was 8.9% in 2023, down slightly from 9.3% in 2019 but still substantial in absolute terms (Audit Analytics).
Goodwill impairment activity in 2024:
| Metric | 2024 figure | Year-over-year change |
|---|---|---|
| Total U.S. goodwill impairments recorded | $96 billion | +16% from $83B in 2023 |
| Number of impairment events | 273 | Across 8,134 monitored companies |
| Top 10 companies' share of total impairments | 53% | Heavily concentrated at large-cap |
| S&P 500 goodwill on balance sheets | ~$4 trillion | Near five-year high |
| Goodwill as % of S&P 500 total assets | 8.9% | Down from 9.3% in 2019 |
Source: Kroll 2025 U.S. Goodwill Impairment Study; Audit Analytics; Calcbench
Goodwill concentration by deal type is also relevant. In technology and software acquisitions, goodwill typically represents 60-75% of the total purchase price, because the acquired value sits in the workforce, code base, and customer relationships rather than tangible assets. In consumer goods and healthcare, goodwill runs 35-45% of purchase price. Industrial and energy deals typically carry 15-25% (CT Acquisitions 2025).
That concentration matters for impairment exposure. Tech acquirers carry the most goodwill relative to what they paid, and are most exposed when market multiples or projected revenues contract.
2. What goodwill impairment testing involves (the workload AI targets)
Understanding what AI actually automates requires understanding the testing workflow. The process has two regulatory frameworks:
ASC 350-20 (U.S. GAAP): Single-step test. Compare the reporting unit's carrying amount to its fair value. If carrying value exceeds fair value, that excess is the impairment loss. Impairment reversals are prohibited. Interim testing is required whenever a triggering event (stock price decline, loss of key customer, macro shock) suggests impairment may have occurred.
IAS 36 (IFRS): Annual mandatory quantitative test applied to all goodwill-bearing Cash-Generating Units (CGUs), with no qualitative skip option. Uses a recoverable amount comparison. Uniquely permits impairment reversals in some cases.
The manual workflow for a large goodwill balance:
| Stage | Manual time | AI potential |
|---|---|---|
| Data gathering and normalization | 8-16 hours | High automation potential |
| CGU/reporting unit identification | 4-8 hours | Moderate (judgment required) |
| DCF model construction | 4-8 hours per model | High (AI builds in under 10 minutes) |
| WACC and benchmarking | 2-4 hours | High (automated range check) |
| Sensitivity analysis | 3-6 hours | High (AI runs scenarios in parallel) |
| Documentation and disclosure | 4-8 hours | Moderate (AI drafts, human reviews) |
| Auditor support and review | 15-30 hours | Moderate (audit trail automation) |
| Total | 60-100 hours | Compression potential: 50-70%+ |
Source: CORAA.ai 2025; Sofer Advisors 2025; Eton Valuation 2025
External valuation fees for a professional goodwill impairment study run $7,500-$25,000+ per reporting unit engagement. Simple single-unit assessments fall at the lower end. Multi-unit analyses involving complex allocation methodologies, litigation support, or expert testimony reach or exceed the upper range (Eton Valuation; Sofer Advisors 2025).
For a Fortune 500 company managing goodwill across multiple reporting units, the full annual testing cycle - internal labor plus external fees - can run $500,000 to over $2 million in combined cost.
3. AI adoption rates in finance and impairment workflows
Finance AI adoption has grown substantially but remains uneven. Most CFOs describe AI as a priority; far fewer have it running across their finance function.
- 59% of finance leaders report using AI in their finance function in 2025, up from 37% in 2023 and 58% in 2024 (Gartner November 2025 survey of finance executives)
- Only 11% of CFOs say AI is fully operational in their finance functions; 35% are in pilots or proofs of concept (L.E.K. 2025 Office of the CFO Survey, approximately 100 CFOs)
- 87% of CFOs at companies with $1B+ in revenue expect AI to be "extremely or very important" to finance operations in 2026, up from prior years (Deloitte Q4 2025 CFO Signals Survey, 200 CFOs)
- 70% of finance professionals plan to invest in AI within five years; over two-thirds are still in the exploratory stage (Wolters Kluwer October 2024 survey, 181 finance professionals globally)
- Among those who have already deployed AI, 60% describe the outcome as successful (Wolters Kluwer 2024)
Adoption specific to audit and impairment procedures:
- AI testing adoption in audit and assurance contexts grew from 7% in 2023 to 16% in 2025 (aggregated industry data)
- Goodwill impairment is consistently among the most frequently cited Critical Audit Matters (CAMs). As of the most recent Audit Analytics data, 22% of all CAMs address asset impairment and recoverability
- Goodwill-only CAMs numbered 18 in the analysis period; combined goodwill and intangible impairment CAMs numbered 28 (Audit Analytics)
The gap between aspiration and deployment reflects the complexity of goodwill testing specifically. Unlike accounts payable or expense report processing, impairment testing is relatively low-frequency (annual, plus interim triggering events), highly judgment-intensive, and carries restatement risk if assumptions are materially wrong. Those characteristics make CFOs and controllers cautious about automating the process entirely - AI is entering as a tool for the mechanical stages, not as a replacement for the valuation judgment.
4. Time and cost impact of AI goodwill impairment automation
The most concrete data on AI's impact in impairment workflows comes from DCF modeling efficiency, which is the highest-time-cost mechanical stage.
AI automates revenue projection, WACC calculation, terminal value estimation, and sensitivity analysis. Industry data and vendor benchmarks show AI-generated DCF models completing in under 10 minutes versus 4-8 hours for manual construction (CORAA.ai; TheBricks.com 2025). That is a 96-99% reduction in modeling time per reporting unit.
Audit and close-cycle cost data from adjacent automation shows consistent patterns:
- Finance teams automating document-intensive processes report 70-80% less processing time per item (aggregated 2024-2025 data)
- 82% of accounting departments using AI have seen error reductions of 15-25 per 1,000 transactions (receiptsai.com, June 2026 aggregation)
- AI audit accuracy improvement: 92% improvement cited across sampled transactions; error reduction of 78% (receiptsai.com 2026)
- External valuation fee reduction from AI pre-processing: estimated 30-50% per engagement, as AI handles data normalization and initial model construction that previously required billable valuation hours (Eton Valuation; Sofer Advisors 2025)
CFO-level cost data reinforces why the investment case is moving:
- 49% of CFOs named "automating processes to free employees for higher-value work" as their top finance talent priority (Deloitte Q4 2025 CFO Signals)
- 53% of CFOs called automation and technology upgrades the most proven cost-control lever available to them (Deloitte Q4 2025)
- Gartner projects that CFOs implementing strategic AI will unlock an additional 10 margin points of growth by 2029 (Gartner April 2026, based on survey of 314 organizations conducted September-October 2025)
5. Accuracy and error reduction in AI-assisted impairment testing
Goodwill impairment testing is sensitive to assumption errors. A WACC that is 100 basis points too low or too high can shift the fair value conclusion enough to change whether an impairment is recorded. Manual DCF construction carries specific error risks: broken formula references, stale comparable benchmarks, inconsistent terminal value assumptions across reporting units, and transcription errors when populating models from source documents.
AI tools address these risks differently depending on the stage:
| Error type | Manual risk | AI impact |
|---|---|---|
| Calculation/formula errors | High (spreadsheet formulas) | Near-eliminated with model automation |
| Stale benchmark data | Moderate-high | Automated benchmark refresh |
| Inconsistent assumptions across units | Moderate | AI applies consistent logic |
| Aggressive growth rate assumptions | Moderate (confirmation bias) | AI flags vs. industry ranges |
| Useful-life or attrition estimate errors | Moderate | Human-intensive; AI supports, not replaces |
| Triggering event identification | High (judgment-based) | Low AI contribution |
| Disclosure drafting errors | Moderate | AI drafts; human reviews for accuracy |
Supporting accuracy data from adjacent accounting automation:
- Tax AI tools achieve 98% compliance accuracy versus 85% for manual processes (receiptsai.com 2026)
- OCR and data extraction from financial documents: 97% accuracy rate as of 2024
- AI delivers a 45% increase in anomaly detection rates; fraud detection using AI shows 50-60% reduction in false positives versus rule-based systems
The UK's Financial Reporting Council flagged in June 2025 that the Big 4 had embedded AI into audit processes without formally measuring the impact on audit quality - a finding that applies directly to goodwill impairment work (Accountancy Age, citing FRC June 2025 report). The PCAOB has identified ongoing deficiencies specifically related to business combinations, long-lived assets, and auditor determination of CAMs. Neither regulator has restricted AI use in audit workflows, but both are signaling that governance around AI-assisted audit judgment needs formal measurement.
6. How AI changes the work of valuation teams and CFOs
AI goodwill impairment automation does not reduce headcount in valuation or accounting teams in the near term. It changes where those people spend their time.
Under manual processes, a valuation analyst on a goodwill impairment engagement spends the majority of billable hours on mechanical stages: gathering financials, building the DCF model, running sensitivity tables, formatting documentation. Senior analyst and manager time goes to reviewing those mechanics and checking assumptions.
When AI handles the DCF build, benchmarking, and initial sensitivity runs, the time distribution changes:
- Analysts concentrate on data quality, source document review, and exception handling rather than model construction
- Senior staff spend more time on WACC assumption validation, CGU identification, triggering event assessment, and auditor coordination
- Partners and CFOs spend more time on disclosure language and audit negotiation
This restructuring of labor toward judgment-intensive work is the consistent pattern in AI finance automation broadly. See our AI and human workers side by side collaboration statistics 2026 for data across other finance functions.
By 2029, Gartner projects that 40% of FP&A teams at large enterprises will use AI-enabled simulation tools to replace bottom-up manual planning - up from 5% today (Gartner 2026). Goodwill impairment scenario modeling fits squarely in that category.
7. How the Big 4 are deploying AI for audit and impairment work
The major accounting firms are investing substantially in AI for audit functions, which includes impairment testing procedures. Their deployment models show what AI-assisted goodwill testing looks like in practice:
| Firm | Platform | Reported deployment |
|---|---|---|
| Deloitte | Omnia; Zora AI (with Nvidia); Anthropic Claude | Anthropic Claude deployed to 470,000 employees; generative AI for risk assessment and financial analysis |
| EY | EY.ai / Helix | 150 AI agents deployed to 80,000 tax and audit staff; scaling to 100,000 agents by 2028; automated document review, compliance, and data collection |
| KPMG | Clara / Ignite; Microsoft partnership | $2 billion committed to Microsoft AI/cloud; targeting $12 billion in AI-enabled revenue; AI integrated into core audit procedures |
| PwC | GL.ai; H2O.ai integration | AI agents for tax research, compliance review, financial statement analysis |
Sources: ChatFin 2026; Consultancy.uk 2025; Accountancy Age 2025; Deloitte/Anthropic public announcements
For goodwill specifically, AI-native valuation tools have entered the market. CORAA.ai offers AI-assisted valuation audit procedures specifically for goodwill impairment testing under IAS 36 and Ind AS 36, with automated DCF checks, CGU identification support, and sensitivity analysis. Finrep.ai covers goodwill impairment disclosure automation under ASC 350-20 and IAS 36.
These are layered on top of the Enterprise Performance Management (EPM) platforms that many companies already use for scenario modeling. The EPM software market reached $7 billion globally in 2024, growing 13.7% year-over-year and projected to reach $9.4 billion by 2029 at a 5.9% CAGR (AppsRunTheWorld 2025). Oracle leads with 20.3% market share; the top five vendors - Oracle, SAP, Workday, IBM, and Anaplan - hold 45-50% of EPM revenue.
8. Market size of AI in finance
AI goodwill impairment automation has not been broken out as a standalone market category by any major analyst firm, but the surrounding markets provide context for the investment scale:
| Market segment | 2024 size | Projected size | CAGR |
|---|---|---|---|
| AI in accounting | $4.87-7.52 billion | $68.75-96.69 billion (2031-2033) | 39.6-44.6% |
| AI in finance broadly | $38.36 billion | $190.33 billion (2030) | 30.6% |
| EPM software (includes impairment modeling) | $7 billion | $9.4 billion (2029) | 5.9% |
Sources: Grand View Research 2025; Mordor Intelligence 2026; MarketsandMarkets 2025; AppsRunTheWorld 2025
The wide range in AI accounting market estimates reflects definitional differences: narrower definitions covering AI-native accounting software, versus broader ones covering all AI applications in any finance-adjacent workflow. The directional story - rapid growth across all definitions - is consistent.
9. Audit and disclosure compliance impact
Goodwill impairment draws significant auditor attention because useful-life and fair-value assumptions carry subjectivity, and a materially wrong conclusion creates restatement risk.
Impairment disclosures operate across three levels under ASC 350-20 and IAS 36: financial statement notes (gross goodwill and accumulated impairment by segment), MD&A critical accounting estimates (the judgment and sensitivity behind the conclusion), and interim triggering-event disclosures. Getting any of these wrong - even in timing or completeness - creates SEC comment letter risk.
AI affects audit compliance in two directions: reducing the mechanical errors that generate audit inquiries, and automatically building the documentation trail auditors require.
Specific data:
- 22% of all Critical Audit Matters (CAMs) address asset impairment and recoverability (Audit Analytics)
- Goodwill impairment is consistently among the five most frequently cited CAM topics in PCAOB filings
- AI tools are increasingly used to pre-validate management's impairment models before auditors engage, reducing back-and-forth review cycles
- AI-generated documentation - rollforward schedules, sensitivity support, assumption logs - reduces audit fieldwork time on comparable accounting automation tasks by an average of 27% (Deloitte 2025, cited across financial close automation studies)
- Companies using AI for adjacent close automation (amortization, reconciliation) report 43% fewer auditor queries related to schedule accuracy (PwC Finance Automation Survey 2025)
For related compliance automation data, see AI compliance automation statistics 2026 and AI revenue recognition automation statistics 2026.
Key takeaways
AI goodwill impairment automation statistics for 2026 show a technology that has arrived in the mechanical stages of testing but has not displaced human judgment where it counts.
The mechanical workload - DCF construction, sensitivity analysis, benchmarking, documentation - compresses substantially with AI. Cutting DCF build time from 4-8 hours to under 10 minutes per model is a real change in how valuation teams allocate their hours. For organizations managing goodwill impairment testing across multiple reporting units each year, the cumulative savings on internal labor and external fees add up.
The judgment stages are a different story. Determining whether a triggering event has occurred, setting WACC assumptions, estimating revenue growth for a struggling reporting unit - these require management judgment that AI cannot supply from historical patterns. The PCAOB and FRC are both signaling that AI-assisted audit work needs formal quality measurement, not just deployment.
What changes is where the hours go. Senior finance and valuation professionals spend more time on the assumptions that change the outcome and less time building spreadsheets.
For companies that need hands-on support with period-end close, impairment documentation, and valuation data preparation without building that capacity internally, Stealth Agents virtual assistants include finance and accounting professionals trained in AI-assisted close workflows and valuation support.
For related research, see AI depreciation automation statistics 2026, AI fixed asset management automation statistics 2026, and AI general ledger automation statistics 2026.
Frequently Asked Questions
What do the 2026 AI goodwill impairment automation statistics show?
The data shows AI reducing the mechanical workload in goodwill impairment testing while leaving judgment-intensive stages to human accountants and valuation professionals. DCF build time drops from 4-8 hours to under 10 minutes per model. External valuation fees decrease by an estimated 30-50% when AI pre-processes data and builds initial models. Adoption in audit contexts grew from 7% in 2023 to 16% in 2025 and is continuing to expand.
How does AI goodwill impairment automation affect valuation teams?
AI shifts how valuation analysts spend their time. Mechanical stages - gathering data, building DCF models, running sensitivity tables - compress substantially. Senior staff and partners redirect time toward assumption validation, CGU identification, triggering event assessment, and auditor coordination. The headcount impact is minimal in the near term; the workload composition changes.
What does goodwill impairment testing cost manually versus with AI?
Manual impairment testing for a single reporting unit at a mid-to-large company runs $7,500-$25,000+ in external valuation fees, plus 60-100 hours of internal team time. With AI handling data normalization, DCF construction, and sensitivity runs, external fees decrease by an estimated 30-50% and internal hours compress by a similar margin, though the savings depend heavily on reporting unit count and data readiness.
What are the main risks of AI goodwill impairment automation?
The main risks are in the judgment stages AI cannot reliably handle: WACC assumption setting, CGU fair value interpretation, and triggering event determination. The UK FRC flagged in June 2025 that the Big 4 are deploying AI in audit without formally measuring its impact on audit quality. Governance frameworks for AI-assisted impairment testing are still developing across the profession.
Related Reading
Tags
Ready to put this into practice?
Book a free 15-min match call
Tell us what role you're filling. We'll match you with a pre-vetted virtual assistant - or tell you honestly if we're not the right fit.
Book a free call →