AI時(shí)代的抉擇-以“信”筑基行穩(wěn)“智”遠(yuǎn)-德勤-2025-01-宏觀大勢

DeloitteAccessEconomics1144保護(hù)隱私承擔(dān)責(zé)任4%%%%%%凸4235668重教治理與控制靠可健穩(wěn)妥穩(wěn)全安用戶保護(hù)用戶友好可預(yù)測一致準(zhǔn)確安全私隱護(hù)保可審核可解釋合理一致同意保密釋解可明透無偏見包容公平可訪問可視公監(jiān)平管合中規(guī)立與監(jiān)督注重可持續(xù)性增值人性化公共和社會利益可適應(yīng)負(fù)責(zé)任可解決所有權(quán)問責(zé)擔(dān)技能塑與育德勤高可信重教治理與控制靠可健穩(wěn)妥穩(wěn)全安用戶保護(hù)用戶友好可預(yù)測一致準(zhǔn)確安全私隱護(hù)保可審核可解釋合理一致同意保密釋解可明透無偏見包容公平可訪問可視公監(jiān)平管合中規(guī)立與監(jiān)督注重可持續(xù)性增值人性化公共和社會利益可適應(yīng)負(fù)責(zé)任可解決所有權(quán)問責(zé)擔(dān)技能塑與育德勤高可信人工智能框架TM承責(zé)任發(fā)展中成熟初級發(fā)展中成熟初級程序與控制66%22%77%支柱一%%%%%支柱二支柱三%支柱四支柱五人文關(guān)懷77%%%%%%%%%%%得社會運(yùn)營許可。%%%%%72%%%%%04有效AI治理的一大益處是提高對AI解決方案成效或成2223產(chǎn)和運(yùn)營以及研發(fā)方面使用AI的可能性要高出兩倍附錄三。 12 2425建議一11建議二247%66%2627建議三3328建議四44理益處。受訪者數(shù)量6430 32組織架構(gòu)3334e程序與控制3536初級組織占比成熟組織占比組織架構(gòu)64.823%37員工人數(shù)38標(biāo)準(zhǔn)誤差指數(shù)得分0.305430.06460.3996標(biāo)準(zhǔn)誤差0.004標(biāo)準(zhǔn)誤差指數(shù)得分0.05550.0039決定系數(shù)0.3352調(diào)整后的決定系數(shù)0.2965標(biāo)準(zhǔn)誤差0.004調(diào)整后的決定系數(shù)0.2707標(biāo)準(zhǔn)誤差指數(shù)得分0.00310.0884決定系數(shù)0.0677調(diào)整后的決定系數(shù)0.004769||%%%%40》》%%%%41%%% 42iidd%%%%43%%%%44 》》%%%%45》》%%%%46%%%%47%%%%48%%%%49》》%% %50ff%%%%w技術(shù)實(shí)施挑戰(zhàn)（37%）wAI實(shí)施戰(zhàn)略和愿景缺位（33%）%%%%541Deloitte(2024)“GenerativeAIinAsiaPacific”,/nz/en/services/consulting/perspectives/generative-ai-in-asia-pacific-may-2024.html3DeloitteAccessEconomics(2024)“ACSAustralia’sDigitalPulse2024:DecodingtheDigitalDecade”,/au/en/services/economics/perspectives/acs-australias-digital-pulse-decoding-the-digital-decade.html4IBM(2024),“CostofadatabreachReport”,/reports/data-breach5CapgeminiResearchInstitute(2020),“AIandtheethicalconundrum”Report,https://www.capgempress-releases/ai-and-the-ethical-conundrum-report/6DeloitteCentreforRegulatoryStrategy(2024),“GenerativeAI:ApplicationandRegulationinA/au/en/Industries/financial-services/analysis/generative-ai-application-regulation-asia-pacific.html7Deloitte(2024)StateofAIinEnterprise,/content/dam/Deloitte/us/Documents/consulting/us-state-of-gen-ai-q3.pdf8Deloitte(2024)StateofAIinEnterprise,/content/dam/Deloitte/us/Documents/consulting/us-state-of-gen-ai-q3.pdf9Haresamdramet.al,IEEEAccess(2023),“ThreelevelsofAItransparency”,httdocument/1004210911Deloitte(2023),Dataprivacyandsecurityworriesareontherise,whiletrustisdown.Consumerdataprivacyandsecurity|DeloitteInsights12DeloitteAccessEconomics(2023),“GenerationAI:readyornot,herewecome!”,/co