从合规到信心：拥抱新思维，提升负责任人工智能成熟度

Embracing a new mindset to advance responsible AI maturity Contents Page 05 Page 03 Page 10 From compliance tovalue: companies areacknowledging the impactof responsible AI Risky business:the three challenges ofthe current risk landscape Executive summary:rethinking responsible AI Page 15 Page 22 Page 12 Responsibility reality check:how ready are companiesfor responsible AI? Becoming reinventionready: the milestones ofresponsible AI maturity Ready, set, grow:five priorities forresponsible AI Executive summary:rethinking responsible AI Responsible AI 101 What is responsible AI?Responsible AI implies taking intentional actions to design, deployand use AI to create value and build trust by protecting against thepotential risks of AI. As generative AI becomes pervasivein business and society, so too do therisks that come with using it. Considerthe chatbot that gave incorrect adviceto a customer, creating liability for thecompany that launched the chatbot. Orthe employee who accidentally exposedproprietary company data after usingChatGPT. Or the algorithm that wronglyflagged thousands of individuals for fraud. In this new environment, the responsibledevelopment and use of data and AIbecomes a critical enabler for organizationsto both minimize the risks that come withusing these technologies and to unlock itsmany benefits—from helping individualsperform tasks better and faster to allowingcompanies to reinvent themselves tocreate more value and gain a competitiveedge through innovation. What is mature responsible AI?Having fully operationalized responsible AI efforts as a platformto take a more systemic, future-orientated approach that unlocksthe true value of AI. The North Star for responsible AI maturity is tobecome a pioneer. No companies are yet at this stage. How do you build a responsible AI program?Our research and work advising clients has shown that allcompanies can benefit from focusing on these five priorities toimprove their maturity and begin to reap the benefits of AI. This is, of course, easier said than done,for the challenges of scaling responsibleAI across an organization are daunting.AI-related risks continue to accumulate asgenerative AI creates and accelerates bothdata and AI risks. New AI-focused laws andregulations are germinating everywhere.And AI value chains1are growing incomplexity, especially as more companiesbecome both developers and buyers ofAI models. 02Conduct AI riskassessments 03Systemic enablementfor responsible AItesting 01 Establish AIgovernanceand principles 05 04 Workforce impact,sustainability,privacy, security Ongoingmonitoring andcompliance Our survey respondents estimate that whena company becomes a pioneer in responsibleAI, its AI-related revenue will increase by18 Responsible data is an integral partof the responsible AI journey It’s important to note that when we talk about responsible AI, the responsibleuse of data is a critical part of the equation. The importance of data in theage of generative AI cannot be underestimated. Most large organizationsare still grappling with longstanding issues around data quality, availabilityand governance. In a separate global survey by Accenture of 2,000 businessleaders in 2024, 48% said their organizations lacked enough high-quality datato operationalize their generative AI initiatives.3Companies with high datareadiness have the right data, with the right quality and quantity. They have ascaled data management and governance practice in place which allows useof data seamlessly across their processes, ensuring responsible adoption andenabling the economic value of data. %on average. Today’s era of generative AI creates newrequirements when it comes to responsibleAI. To be a leader in responsible AI,companies must pursue an anticipatorymindset, commit to continuousimprovement and extend their focusbeyond their organization to their entirevalue chain and wider AI ecosystem. Wecall this new level of maturity “becoming aresponsible AI pioneer”, and no companiesare yet at this stage. survey C-suite executives across 1,000companies, spanning 19 industries and 22countries.2 Click here to learn more about key actions companies can take to improve theirdata readiness in the era of generative AI. We assessed companies’ maturity inresponsible AI by developing a four-stageframework—the higher the stage, thegreater the progress. We then applied thatframework to analyze the organizationaland operational maturity of the 1,000companies we surveyed. Years of client experience showus thatcompanies can take action to systemicallyoperationalize responsible AI. In thisreport, we will explore the actions that willprovide the organizational infrastructureto become a responsible AI pioneer andunlock the true value of AI. They estimate that when a companybecomes a pioneer in responsible AI,its AI-related revenue will increase by18% on average. However, our researchfound that a great majority of companieswe surveyed are