2024年人工智能治理：框架、最佳实践和政策

AIGovernance Frameworks, Best Practices and Policies in 2024 Foreword In 2024, artificial intelligence (AI) has reached a pivotal momentof widespread adoption and capability. While AI offers immensepotential for productivity gains, operational efficiencies, and newrevenue opportunities, it also brings significant challenges aroundgovernance, ethics, and implications in the workforce. Contents 02ImplementingAI Governance Leaders feel pressured to invest and innovate rapidly in AI, oftenwith a ‘fear of missing out’ (FOMO) driving decision-making overprudent long-term planning. There is a fundamental tensionbetween the urgency tocapitaliseon AI's benefits and the criticalneed for appropriate governance and strict human control. Theurgency to establish proper guidelines is not just a technocraticnecessity, but a societal imperative, ensuring that these powerfultools are used in a responsible manner. Ultimately, success will hinge onorganisationsstriking the rightbalance–harnessing AI's potential while implementing thenecessary guardrails to mitigate risks and hazards. 04Appendix:AdditionalFrameworks and Useful Sources This guide aims to help in that goal by compiling insights frommembers ofWinmark’sC-Suite networks who have shared theirgovernance practices for responsibly deploying AI. Acknowledgments We would like to thank all of theWinmarkmembers whoshared their valuable insights and experiences during thedevelopment of this guide. Our insights and recommendations are positioned to impel individuals andorganisationsto develop and action ethical and effective AI policies andprocedures, aligned to wider regulatory frameworks Core values and principles AlignDevelop a clear code of 01 ethics and principles forthe responsible use of AIwithin the organisation (shared byWinmark’smembers for individuals, businesses,organisationsand governments) Key ConsiderationsandFrameworks ImplementDesign, align and 02 implement AI policies andprocedures acrossoperational scenarios andassess its potential risk ImplementingAI Governance Winmark’smember examples 03How to use this guide InspireDive into a collation of checklists, use cases, andbest practices, shared byWinmarkmembers AI Governance:Member examples 01Key ConsiderationsandFrameworks Core values and principles for AI Governance (1/2) When implementing your AIgovernance program,leveraging establishedframeworks can providevaluable structure andguidance and help ensureyou are adopting recognised,public standards. It alsounderpins transparency andconsistency across yourorganisation when it comesto auditing AI tools andevaluating AI systems. Key considerations shared byWinmark’smembers Principles ✓Establish risk mitigation strategies, including rigorous testing, human oversight, andcontrolled piloting of AI use cases before full deployment. Safety Security ✓Implement robust risk assessment processes to identify potential risks associated withAI systems, such as bias, security vulnerabilities, IP infringement, and environmentalimpacts. Analyse potential misuses and harmful capabilities proactively. ✓Assign clear roles and responsibilities for AI governance, such as data stewards or AIethics officers. Transparency Accountability ✓Establish multiple lines of defence, involving developers, risk evaluators, and independentauditors, to ensure accountability. The following key values andprinciples to guide ethical AIimplementation are sharedby Winmark members. Accuracy ✓Integrate AI governance into existing compliance frameworks, policies (e.g., data ethics,vendor assessment), and operating models. ✓Implement appropriate oversight models based on the required level of humaninvolvement: human-in-the-loop, human-on-the-loop, or human-out-of-the-loop. Core values and principles for AI Governance (2/2) By addressing the followingkey considerations in thissection, C-Suite leaders canestablish comprehensive androbust AI governance policiesthatprioritiseethicalpractices, risk management,accountability, andtransparency. Further, itreinforces the commitmentto the responsibledevelopment anddeployment of AItechnologies within theirorganisations. Key considerations shared byWinmark’smembers Principles ✓Promote AI literacy across the organisation through comprehensive training programstailored to different roles and responsibilities. Education ✓Upskill or reskill employees to work effectively with AI systems and address potential jobdisplacement or role changes. ✓Foster a culture of responsible AI development and deployment through education andalignment with organisational values. ✓Establish mechanisms for transparent communication about the organisation's AIpractices, decision-making processes, and potential impacts. ✓Actively participate in industry collaborations, global forums, and multistakeholderdialogues to stay informed about best practices and contribute expertise.* Engagement ✓Foster partnerships with academic institutions and con