信息发布
    当前位置:首页 > 教育发展基金会 > 信息发布
    学院动态 | 人工智能时代内部审计行动共识(V1.0)
    日期:2026-06-23

    人工智能浪潮正重塑各行各业,内部审计领域也迎来全新变革机遇与全新风险挑战。依托学院主办内部审计高质量发展论坛,我院联合中国商业会计学会汇聚业界、学界顶尖专家,共同研讨凝练形成《人工智能时代内部审计行动共识(V1.0)》双语文稿正式发布。

    众多校友深耕审计、财务、风控、企业治理一线,身处 AI 落地应用最前沿。这份共识立足实务、直击痛点,清晰厘清人机协同边界、审计权责底线与行业发展方向,为大家应对数字化转型、驾驭智能审计工具提供清晰行动指引。诚邀各位校友细读参阅,结合自身工作实践交流思考,一同探索AI赋能内部审计的长远发展之路。


     

    人工智能时代内部审计行动共识

    (V1.0)


    序   言

    人工智能(AI)浪潮正在席卷全球,全面改变人类生产生活方式,也在深刻影响着内部审计,带来史无前例的机遇和前所未有的挑战。为了更好地应对变局,提升内部审计价值创造能力,促进高质量发展,中国商业会计学会、上海国家会计学院基于“内部审计高质量发展论坛”,汇集企业、高校、行业协会等领域专家智慧,形成以下共识。


    一、AI是内部审计的赋能者

    AI能高效自动执行数据处理、文本生成、事件检测等工作,将内部审计师从大量低效、重复的劳动中解放出来,让持续审计、全量覆盖与前瞻洞察成为现实,大幅提升内部审计工作质效和价值。


    二、内部审计必须积极拥抱AI

    内部审计组织必须密切关注AI技术发展,主动变革,利用AI重塑审计流程、作业模式以及组织形态,重新定义内部审计在治理架构中的角色,赋能业务发展和风险管控,为治理层和管理层提供战略支撑。


    三、人机协同成为审计的基本范式

    内部审计师与AI深度协同,各有侧重:内部审计师赋予审计高度与温度——战略判断、治理设计、沟通协调与推动整改;AI赋予审计速度与广度——全量数据实时穿透、持续监控与全面覆盖。碳基智慧与硅基智能融合共生,成为审计的基本范式。

    四、审计责任始终在人

    AI无法独立做出职业判断,无法承担审计责任。内部审计师务必独立、审慎、批判地甄别运用AI产出结果。审计结论最终必须由人做出,审计责任最终由人承担,这是AI时代治理信任的最后防线。


    五、AI的审计应用必须可验证

    内部审计必须坚守审计安全与质量生命线。针对AI幻觉、模型漂移、算法偏见等新风险,应当建立人工主导的验证机制,确保审计过程可回溯、逻辑可解释、结果可验证。


    六、驾驭AI,做AI做不了的事

    理解、使用并审计AI,应成为内部审计师的基本素养。内部审计师更须聚焦于洞察、伦理、沟通与治理等复杂场景,做出AI无法替代的职业判断。


    结     语

    人机协同,碳硅共生,护航组织,点亮未来。


    ******************************************************


    Consensus on Internal Audit Actions in the Age of AI

    (V1.0)


    Preface

    The wave of artificial intelligence (AI) is sweeping the globe, fundamentally transforming the way people live and work, and profoundly reshaping the practice of internal audit — bringing both unprecedented opportunities and challenges. To better navigate this shifting landscape, enhance the value-creating capacity of internal audit, and promote high-quality development, building on the deliberations of the Internal Audit High‑Quality Development Forum, the China Association of Business Accounting and the Shanghai National Accounting Institute have drawn on the collective wisdom of experts across enterprises, universities, and industry associations and reached the following consensus.


    1. AI as an Enabler of Internal Audit

    By efficiently automating tasks such as data processing, text generation, and event detection, AI frees internal auditors from time-consuming, repetitive work. It makes continuous auditing, full-population coverage, and forward-looking insights possible — substantially elevating the quality, efficiency, and value of internal audit.


    2. Internal Audit Must Proactively Embrace AI

    Internal audit functions must closely monitor advances in AI, take the initiative in driving transformation, and leverage AI to reshape audit processes, delivery models, and organizational structures. In doing so, they redefine the role of internal audit within the governance framework, support business development and risk management, and provide strategic value to those charged with governance and management.


    3. Human-AI Collaboration as a New Paradigm for Auditing

    Internal auditors and AI collaborate deeply, each with distinct strengths. Internal auditors bring depth and warmth to audit work through strategic judgment, governance design, communication and coordination, and facilitating corrective actions. AI brings speed and breadth through real-time full-population data analysis, continuous monitoring, and comprehensive coverage. The integration and coexistence of carbon-based wisdom and silicon-based intelligence form the new paradigm for auditing.


    4. Audit Accountability Always Rests with Humans

    AI cannot exercise professional judgment independently or assume audit accountability. Internal auditors must critically assess, validate, and apply AI-generated outputs with independence and due care. Humans must make the final audit conclusions and assume the ultimate audit accountability. This is the last line of defense for trust in governance in the AI era.


    5. AI Applications in Auditing Must Be Verifiable

    Internal audit must steadfastly safeguard the lifeline of audit integrity and quality. In response to emerging risks such as AI hallucinations, model drift, and algorithmic bias, human-led verification mechanisms must be established to ensure that audit processes are traceable, reasoning is explainable, and results are verifiable.


    6. Harness AI; Do What Only Humans Can

    Understanding AI, using AI, and auditing AI systems should become core competencies for every internal auditor. Beyond this, internal auditors must focus on complex, context-dependent areas that require insight, ethics, communication, and governance, and exercise professional judgment that AI cannot replicate.


    Conclusion

    Through human-AI collaboration and carbon-silicon symbiosis, we safeguard organizations and illuminate the path forward.



    ▷ 来源:教务二部

    ✅ 已复制