AICPA and AIUC develop assurance certificate for artificial intelligence systems

The AICPA and the Artificial Intelligence Insurance Company (AIUC) have developed a joint assurance certificate that will provide an independent and comprehensive examination of business risks associated with artificial intelligence systems and agents.

The AICPA and AIUC partnership offers a single CPA-led service combining the AICPA-led System and Organization Controls (SOC 2) report with the AIUC-1 certification. Over the past year, more than 40,000 organizations received a SOC 2 attestation report or an AIUC-1 certification. While SOC 2 reports demonstrate controls for data protection, risk management, and building stakeholder trust, the AIUC-1 certification provides a standard for AI agent security, safety, and reliability.

AICPA President and CEO Mark Koziel stated that companies are making massive investments in artificial intelligence and that investors need a clearer picture of whether the technology is safe, reliable, and managed. AIUC Co-Founder Rajiv Dattani noted that AI agents are rapidly executing real business tasks, and this joint certificate provides independent answers regarding security posture and whether agents are behaving as intended. It was reported that AIUC-1 goes beyond policy reviews to perform live testing across hundreds of risks and attack combinations, such as hallucinations, data leakage, manipulative prompts, and adversarial prompt injection.

It was stated that multiple CPA firms are already offering this assurance certificate to their clients, with a full launch expected in early 2027 following the completion of initial work. To meet security, safety, and reliability standards, SOC 2 reports require annual renewal, whereas the AIUC-1 certification requires quarterly technical retesting in addition to annual renewal.