dnl Expands Into U.S. With AI-Powered GAAP Disclosure Checklist
German audit software provider dnl launched a U.S. GAAP disclosure checklist built on AI agents on November 16, its first product for U.S. auditors.
Germany-based audit software provider dnl launched a disclosure checklist for GAAP reporting in the U.S., built on AI agents, on Monday, November 16, 2026, its first product for the American market.
Aimed at audit firms and the reporting teams that prepare financial statements, the dnl checklist for GAAP in the U.S. matters because disclosure checking is among the most repetitive work in an engagement. The agents read a client’s annual report, test it against each required disclosure and check the numbers, so that the auditor spends time on judgement rather than on cross-referencing.
The launch also adds to a crowded field. Caseware and DataSnipper already sell agent-based disclosure review for GAAP in the U.S., which means the dnl checklist enters the market as a late arrival rather than as the first mover.
Accountio.The go-to weekly newsletter, straight to your inbox.Weekly newsletter, straight to your inbox.
What the dnl U.S. GAAP checklist covers
At launch the dnl software automates the checklist for GAAP in the U.S. for nonpublic companies, from a single-entity business to a private equity-backed group. It also covers not-for-profit reporting under ASC 958 for foundations, associations and other nonprofits, as well as financial reporting for cities, counties and other local governments.
The dnl checklist can run against GAAP in the U.S. and ESG frameworks, or against a firm’s own checklist, and it reviews both financial and sustainability statements. Firms that already keep their own disclosure checklists therefore do not have to abandon them to use the product. Teams preparing for a nonprofit engagement may want to compare the output with their own nonprofit audit and Form 990 pack.
How the AI agents check disclosures and figures
The agents read the company’s annual report, check every required disclosure and verify that the numbers are accurate and consistent. Where an agent is confident in an answer, it completes the item, and where it is not, it hands the question to the auditor as a proposal.
A second set of checks covers the figures themselves. The software tests totals, cross-references, notes and prior-year comparatives across PDF, Excel and report tables, and it flags only the numbers that do not agree. When a client sends a new draft, completed work carries over, and the software shows what changed between the two drafts and whether the change matters.
The product also supports work before the audit starts. During drafting, a report-review agent checks each finished report against the financial statements and the firm’s own reporting standards before it reaches internal review, explains each comment and highlights the exact words that triggered it. Everything the agents and the team find, from a flagged number to a checklist gap, lands in one list to raise with the client, and it exports as one package.
What auditors still sign off on
The agents produce a checked result in which every answer links to the page it relies on, and the auditor reviews, approves and signs. Andreas Schindler, co-founder and chief executive of dnl, put the division of labour this way in a statement on Monday.
“U.S. auditors face the same pressure we’ve seen in every market we serve: more disclosure requirements, tighter deadlines and no room for error.”
Andreas Schindler, Co-Founder and CEO, dnl
Schindler added that the checking and cross-referencing links every answer to its evidence and leaves the judgment with the auditor. The preparer accepts or adjusts each answer, so a person remains responsible for every line on the checklist.
The company’s website addresses the professional standards directly. Under AU-C 500, as amended by SAS 142, the output of an automated tool is information the auditor assesses for relevance and reliability before using it as evidence, and dnl’s page links each answer to its source so that this evaluation is part of the work.
For documentation, AU-C 230 requires the file to show who performed the work and when, and who reviewed it and when. dnl records who accepted or adjusted each answer and who reviewed it, with the date, for every question on the checklist.
That design should matter to inspectors as much as to engagement teams, because the weakness of many AI tools is an answer that cannot be traced. Firms that adopt any such tool will still want a defined software implementation checklist and a clear point at which reviewers test the agent’s output.
dnl’s footprint beyond the U.S. GAAP checklist
The U.S. launch builds on a product range that dnl already sells in Europe and beyond. Its website lists AI-driven disclosure checklists covering IFRS, U.S. GAAP and the European Sustainability Reporting Standards (ESRS), alongside a numbers product that runs tick-and-tie and consistency checks across internal, external and prior-year figures. A third product, which maps trial balance lines to balance sheet and income statement captions, is listed as coming soon.
dnl reports 140 firms, 15,000 auditors and 60,000 completed audits on its platform. Its site shows the logos of KPMG, BDO, RSM, Baker Tilly, Forvis Mazars, Moore, HLB and Nexia, as well as Porsche, and carries testimonials from firms in Germany, the Czech Republic and Australia. Schindler described it as a platform “that thousands of auditors already trust”.
The company’s site says its checklist agents auto-populate responses to complete 90% of validations in real time. That figure comes from dnl and has not been tested against a U.S. engagement, which is where the product has yet to build a record.
Security, data hosting and availability
The checklist has held ISO/IEC 27001:2022 certification since 2024, with annual surveillance audits. For U.S. buyers of the dnl GAAP checklist, the more practical point is that U.S. client data stays in the U.S., hosted in the United States, and is encrypted with AES-256 at rest and TLS 1.2 or higher in transit, with strict isolation between firms.
Data residency often decides whether a firm’s information security committee approves a tool, particularly for firms that hold a written information security plan and must justify each vendor that touches client files. dnl has not published a price for the U.S. product, and the checklist is available now for the three client types listed above.
The competition and what comes next
Where dnl differs from the U.S. vendors is its scope across frameworks. A firm that audits both U.S. GAAP and IFRS clients, or that must report on sustainability, can run all three kinds of review in one tool, whereas several rivals started with a single framework. Larger firms are building similar capability in-house, as the Deloitte Omnia network of AI agents shows, which leaves independent vendors competing on how well they fit smaller and mid-tier practices.
The first test for dnl in the U.S. will be public company work. The launch covers nonpublic companies, nonprofits and local governments, so SEC registrants fall outside it for now. The company’s next step is to show U.S. audit firms that the work it saves is real, which will depend on early engagements through the 2026 year-end audit season.
Accountio.
The go-to weekly newsletter for accounting professionals. Trusted by 10,000+ industry leaders to deliver the tech trends and insights that matter most. Join them today.
