人工智能治理专业实践报告2025

Executive summaryPart I. Building an AI governance programCase study: AI governance at MastercardCase study: AI governance at TELUSCase study: AI governance at Boston Consulting GroupPart II. Professionalizing AI governanceCase study: AI governance at KrollCase study: AI governance at IBMPart III. Leadership and accountabilityCase study: AI governance at RandstadCase study: AI governance at CohereConclusionContacts Artificial intelligence governance can provide businesseswith the certainty they need to continue to innovate with AI atscale, building faster, better, more reliable products that aretrusted by both consumers and enterprise partners alike. AI enables global businesses to compete. It makes themfaster, more efficient and more competitive. In order to adoptAI confidently, businesses need certainty. The field of workwe are in focuses on how to give enterprises the certaintythat their AI systems are accountable, trustworthy and safe,removing the barriers to their AI adoption so they can competewith enterprises that are already using AI to win business. Enterprises have adopted AI, and to stay competitive andcontinue using it, they now need to manage the risks at scale.Managing AIrisk has truly become a reality for enterprisesthat must ensure compliance with hard regulations that havealready come into force, such as the EU AI Act, as well asadditional regulations that are already emerging in 2025, suchas South Korea's AI Act. The first penalties for noncompliancewith AI-specific laws will begin to set a global precedent,forcing businesses to prioritize governance or face steepconsequences. We believe 2025 will mark the year when AIgovernance becomes a strategic differentiator forcompanies, and we expect to see real commitmentsand actions to manage AI risks - learning fromboth cybersecurity and privacy risk managementbest practices but incorporating new and uniqueapproaches to AI-specific risks. compliance are just some of the domains anddisciplines in the community learning and workingtogether on AI governance. This burgeoning village of professionals workingon AI governance is filling the urgent gap betweenthe demand for experts to implement responsibleAIpractices and the professionals who are readyto do so. As these practices bed in and scale,organizations are reckoning with the businessand strategic benefits of having invested in andprioritized AIgovernance. Organizations navigating the complexity ofAIdevelopment and deployment are increasinglyconsidering how AI governance can be an enablingfunction and success factor for their strategicobjectives on AI. Accordingly, how to design andbuild an AI governance program has fast emerged asa top strategic priority fororganizations. Adopting privacy by design and security by designhas always enabled businesses to act faster and bemore strategic in the market. Responsible AI bydesign is no different; it will be essential in an erawhere businesses that are unable to adopt AI areleft behind. Smart investments in secure and robustAIframeworks will allow global companies to movefaster and scale better in the longer term. This report charts how the pillars ofprofessionalization – the people, their skills andtraining, the tools, and processes – are supportingthe strategic functions and organizational structuresfor AI governance. The report dives into the varietyof ways in which organizations are approaching,designing, and implementing AI governance, andidentifies key and common themes. One such theme is that it takes a village to build anAI governance program. In addition to an uptickin dedicated AI governance roles being created,communities, committees and cross-functional teamsand taskforces are leveraging existing resourcesand structures as well as taking on the necessaryadditional and AI governance-specific upskillingand tooling. Privacy, cybersecurity, informationtechnology, ethics, product, marketing, legal and The promulgation of AI governance legislation, regulationsand standards combined with increasingly complex anddemanding sociotechnical pressures have organizationsprioritizing the building and implementation of AIgovernanceprograms. This report, and the data within it, profiles the extent to whichorganizations are implementing AI governance programs,and how they are doing so. Indeed, survey data shows howthe development and deployment of AI by organizations veryoftengoes hand in hand with AI governance. "governance first" Of surveyed organizations, 77% are currentlyworking on AIgovernance, with a jump tonear 90% for those organizations already usingAI. Importantly, 30% of organizations not yetusing AI reported working on AI governance,perhaps revealing a prevailing "governance first"prioritization of ensuring good governance is inplace before AI use. This is supported by some ofthe case studies, which indicate organizations areimplementing formal AI governance programsafter using AI for smaller use cases but beforeembracing AI as a strat