BlackLine Earns AIUC-1 Certification for Agentic Finance Platform

BlackLine’s Verity AI agents passed nearly 3,000 adversarial tests with zero critical findings to earn AIUC-1 certification, the company announced on November 18.

BlackLine Earns AIUC-1 Certification for Agentic Finance Platform

BlackLine said on November 18, 2026 that AI agents powered by its Verity AI passed nearly 3,000 adversarial evaluations with zero critical findings, earning AIUC-1 certification, a standard written specifically for enterprise AI agents.

For controllers and CFOs, the certificate gives them something to put in front of an audit committee. Finance teams are deploying AI agents across record-to-report, invoice-to-cash and intercompany work, and committees increasingly ask for independent evidence of how those agents are tested and governed, rather than a vendor’s description of its own controls.

The Woodland Hills, California-based company said it is the first agentic financial operations platform to earn the certification. Other AI vendors already hold it, including UiPath, Cursor, Harvey and Fieldguide, but none of them sells close or receivables software, so the claim is about the finance category rather than about AIUC-1 as a whole.

Accountio.The go-to weekly newsletter, straight to your inbox.Weekly newsletter, straight to your inbox.

What AIUC-1 certifies for BlackLine Verity AI

AIUC-1 is maintained by the Artificial Intelligence Underwriting Company, which describes it as the technical certification for agent security. The standard covers six risk areas: data and privacy, security, safety, reliability, accountability and society. Those areas range from prompt injection and unauthorised actions to hallucinations, incident response and logging.

The process has four stages. The Artificial Intelligence Underwriting Company defines the scope, then runs red-team evaluations, typically 1,000 to 5,000 scenarios, followed by an audit of policy, operational and technical controls by an accredited assessor. A certificate and full audit report are then issued, valid for one year, with technical retests each quarter.

The assessor in this case was Schellman, which was named the first accredited auditor for AIUC-1. The standard’s controls are aligned with the ISO 42001 standard for AI management systems, the NIST AI Risk Management Framework and the OWASP Top 10 for Large Language Model Applications.

How BlackLine Verity AI agents were tested

The examination put the Verity AI agents through nearly 3,000 distinct adversarial evaluations across four areas. Prompt injection testing checked whether malicious prompts could override financial rules, and data isolation testing confirmed there was no leakage between tenants or sessions.

Workflow reliability testing assessed tool-calling accuracy, hallucination prevention and strict adherence to accounting logic, while a fourth area reviewed BlackLine’s policies, agent guardrails and supervision protocols. The agents completed every evaluation with zero critical findings.

The company chose to publish a headline benchmark from the certifying body, which gives the tests some context. Across more than 60,000 enterprise evaluations, AI agents hallucinate in 9% of interactions and trigger incorrect tool actions 8% of the time. In accounting, one wrong action can become a misstatement, which is why those rates matter more here than in a customer-service setting.

“We do not expect our customers to rely on self-attested claims about how our AI agents operate. We engaged independent auditors to subject them to rigorous adversarial testing, and we are committed to maintaining that cadence every quarter.”

Owen Ryan, Chief Executive, BlackLine

Ryan added that the standard is one the Office of the CFO should require from every technology partner operating with core financial systems.

Where BlackLine Verity AI fits among finance rivals

BlackLine launched Verity in September 2025 as AI built for the Office of the CFO, and has since added agents to its close products, most recently Verity Prepare for its hardest close process. The vendor’s platform, which Accountio has profiled as a $4 billion cloud business, competes with FloQast and Numeric, which are compared in our month end close software guide.

The new certificate adds to an earlier governance credential, as BlackLine already held ISO 42001 certification for its AI management system. The two cover different ground, because ISO 42001 certifies how an organisation manages AI, while AIUC-1 tests how specific agents behave under attack.

That distinction matters in a market where nearly every vendor now sells agents. Buyers comparing financial close software have mostly had to rely on demonstrations and questionnaires, so a third-party test report with quarterly retesting is a more concrete basis for comparison.

What the BlackLine Verity AI certificate means for CFOs

It covers the vendor’s agents as tested, not how a given company configures them, which permissions it grants them or how its reviewers sign off their output. Controllers still need to decide which postings an agent may prepare and which a person must approve.

Finance leaders can use the report as a procurement document, however.

What to watch after the BlackLine Verity AI audit

The certificate is valid for one year, so BlackLine will need to pass the next quarterly technical test before then, and its first retest result is the next date to watch. A second finding to look for is whether competitors in close and receivables software seek the same certificate, which would turn a point of difference into a baseline requirement. For readers weighing the wider shift, our guide to agentic AI in accounting sets out the risks and the controls firms should expect.