The American Institute of Certified Public Accountants (AICPA) has partnered with Artificial Intelligence Underwriting Company (AIUC) to introduce a new assurance certification for artificial intelligence systems and agents, with PwC confirming its involvement in developing the initiative.
The joint offering, announced on 19 October, combines two existing assurance frameworks into a single engagement led by a certified public accountant (CPA).
The first component, SOC 2, examines an organisation’s controls over areas including security, availability, processing integrity, confidentiality and privacy. The second, AIUC-1, evaluates the safety, security and reliability of AI agents through technical testing.
Unlike traditional control assessments, AIUC-1 includes testing how AI agents behave under potentially harmful conditions, including attempts to manipulate their instructions, extract confidential information or induce unreliable outputs.
According to the AICPA, testing can cover thousands of scenarios, with annual renewal required for both components and quarterly technical retesting under AIUC-1.
Several CPA firms are already delivering the combined offering as part of initial engagements. A wider launch is expected in early 2027.
PwC and KPMG involvement
In a statement published on 19 October, PwC US said it was a founding member of the AICPA working group behind the initiative and an early member of the AIUC-1 consortium.
The firm said it had helped shape the combined report and associated guidance for practitioners.
KPMG also has an existing relationship with AIUC. In August, the firm became the first of the Big Four to obtain AIUC-1 certification for its own aIQ Capture platform, following more than 900 technical tests.
KPMG subsequently joined the AIUC-1 consortium. The certification concerned KPMG’s own technology rather than its participation as an assurance provider in the newly announced joint offering.
A new assurance market
The initiative potentially creates a new assurance market for accounting firms as companies increasingly deploy AI agents capable of accessing sensitive information, interacting with business systems and performing tasks autonomously.
It also raises questions about how professional liability, auditor independence and the boundaries of an assurance opinion will be addressed when firms examine technologies whose behaviour can change after certification.
The development extends the accountancy profession’s assurance activities beyond conventional financial reporting and internal controls into the assessment of AI systems themselves.
Trying to keep up with technology at Deloitte, PwC, EY and KPMG? Start here →
https://www.big4news.com/t/technology-and-ai
Check out the News section and subscribe to Big4News for weekly deep dives and briefings.
About the author: Claudine Cassar is the founder and editor of Big4News, covering audit, consulting, regulation and governance across Deloitte, PwC, EY and KPMG.