《德勤:2025企业生成式人工智能应用现状报告:企业GenAI采用情况追踪(第4份)(英文版)(46页).pdf》由会员分享,可在线阅读,更多相关《德勤:2025企业生成式人工智能应用现状报告:企业GenAI采用情况追踪(第4份)(英文版)(46页).pdf(46页珍藏版)》请在三个皮匠报告上搜索。
1、Deloittes State of Generative AI in the Enterprise Quarter four reportJanuary 2025Now decides next:Generating a new 2IntroductionKey findingsLooking back at 2024Now:Where we areNext:Looking aheadConsiderationsCase studiesAuthorship&AcknowledgmentsAbout the Deloitte AI Institute About the Deloitte Ce
2、nter for Integrated Research About the Deloitte Center for Technology,Media&TelecommunicationsMethodologyTable of contentsIntroductionForewordIt was only about 10 years ago when visionary tech leaders started talking about a future powered by ubiquitous computing and ambient intelligence.Back then i
3、t sounded like science fiction.Today,its real.No where is this future more evident than in the rapid advancement and adoption of AI technologies.New models and tools are gaining greater and greater capabilities and performing more complex reasoning.Even what was state of the art a few years ago pale
4、s in comparison to what we have today.In this AI era,many now believe that Moores Law is effectively dead.And we have every reason to believe that the AI flywheel will continue to accelerate with every week and yearoften referenced as the greatest secular shift of this quarter century.Despite the te
5、chnologys rapid pace,I hear from clients and business leaders who are wondering when it will meet their transformational expectationswhen will business leaders see the value and innovation that has been promised?Just like the internet,cloud,or even mobile,the transformational opportunities werent un
6、covered overnight.But as they became pervasive,they drove significant disruption to business and technology capabilities,and also triggered many new business models,new products and services,new partnerships,and new ways of working and countless other innovations that led to the next wave across ind
7、ustries.As we have experienced the half-life of these waves continues to be shorter.As such,it requires enterprises to be a lot more structurally agile to adapt,embrace and innovate to stay relevant and differentiated.In the following report,we see that most companies are transforming at the speed o
8、f organizational change,not at the speed of technology.This is not surprising but is something that will need to be addressed.That said,many are also already using GenAI to create business value that exceeds their expectationswith compelling new use cases emerging every day.So,what do I say to clien
9、ts who are in the trenches of this transformation?Dont lose focus.Stay curious,and challenge the orthodoxies of your organizations.GenAI and AI broadly is our realityits not going away.While there are more questions than answers,but to stay in the game,leaders must be willing to try,do unconventiona
10、l things,learn and help mature.State of GenAI in the Enterprise is a snapshot in time of this great transformation.An opportunity for you to see where and how organizations across industries are finding their way.I hope it serves to spark new ideas and new approaches that help illuminate the path to
11、 your organizations AI-fueled future.Ranjit Bawa,Principal,US Chief Strategy and Technology Officer3IntroductionGenerating a new futureFor the past year,Deloitte has been conducting quarterly global survey reports and executive interviews focused on Generative AI(GenAI)in the enterprise.We titled ou
12、r study Now decides next because we believed in GenAIs potential to dramatically transform how businesses operateand that the actions companies take today will have a decisive impact on their ability to succeed with GenAI in the future.And thats exactly what we found.As with previous transformationa
13、l technologies,the initial excitement and hype about GenAI has gradually given way to a mindset of positive pragmatism.Many companies are already seeing encouraging returns on their early GenAI investments.However,those companies and others have learned that creating value with GenAIand deploying it
14、 at scaleis hard work.Although the technology at times seems like magic,there is no magic wand when it comes to GenAI adoption,deployment,integration and value creation.4 4There is a speed limit.GenAI technology continues to advance at incredible speed.However,most organizations are moving at the sp
15、eed of organizations,not at the speed of technology.No matter how quickly the technology advancesor how hard the companies producing GenAI technology pushorganizational change in an enterprise can only happen so fast.Barriers are evolving.Significant barriers to scaling and value creation are still
16、widespread across key areas.And,over the past year regulatory uncertainty and risk management have risen in organizations lists of concerns to address.Also,levels of trust in GenAI are still moderate for the majority of organizations.Even so,with increased customization and accuracy of modelscombine
17、d with a focus on better governanceadoption of GenAI is becoming more established.Some uses are outpacing others.Application of GenAI is further along in some business areas than in others in terms of integration,return on investment(ROI)and expectations.The IT function is most mature;cybersecurity,
18、operations,marketing and customer service are also showing strong adoption and results.Organizations reporting higher ROI for their most scaled initiatives are broadly further along in their GenAI journeys.IntroductionKey findingsAll statistics noted in this report and its graphics are derived from
19、Deloittes fourth quarterly survey,conducted July September 2024;The State of Generative AI in the Enterprise:Now decides next,a report series.N(Total leader survey responses)=2,773.Percentages in this report and its charts may not add up to 100,due to rounding.Generative AI is an evolving area of ar
20、tificial intelligence and refers to AI that in response to a querya promptcan create new text,images,video and other assets.Generative AI systems can interact with humans and are builtor“trained”on datasets that range in size and quality from small language models(SLMs)to large language models(LLMs)
21、.Generative AI is also referred to as“GenAI.”Evolving upon GenAI technologies,emerging AI agents are software systems that can complete complex tasks and meet objectives with little or no human intervention.They are called“agents”because they have the agency to act independently,planning and executi
22、ng actions to achieve a specified goal.Related,the vision for agentic AI is that autonomous AI agents will be able to execute assigned tasks consistently and reliably by acquiring and processing multimodal data,using various tools to complete tasks,and coordinating with other AI agentsall while reme
23、mbering what theyve done in the past and learning from their experience.5The focus is on core business value.A strategic shift is emerging,from technology catch-up to competitive differentiation with GenAI.Beyond the IT function,organizations tend to focus their deepest GenAI deployments on parts of
24、 the business uniquely critical to success in their industries.The C-suite sees things differently.Relative to leaders outside of the C-suite,CxOs tend to express a rosier view of their organizations GenAI investmentsand how easily and quickly GenAIs barriers will be addressed and value achieved.Its
25、 critical that CxOs move on from being cheerleaders to being champions for achieving organizational efficiency and market competitiveness.Agentic AI is here.Agentic AI is gaining interest as a breakthrough innovation that could unlock the full potential of GenAI,with GenAI-powered systems having the
26、“agency”to orchestrate complex workflows,coordinate tasks with other agents,and execute tasks without human involvement.However,agentic AI is not a silver bullet and all the broad challenges currently facing GenAI still apply.IntroductionKey findings6Our previous quarterly report said the clock was
27、ticking to prove valueand this remains true today.Senior decision-makers might not be demanding tangible value and financial results from GenAI yet,but they soon will be.More and more organizations are moving from GenAI experimentation to deployment and scalingwith proven use cases emerging and sign
28、ificant ROI being achieved through the most advanced GenAI initiatives.Whats more,despite some feelings of disillusionment and unmet expectations,the vast majority of organizations we surveyed are taking a realistic perspective and showing sustained commitment in their quest for value from GenAI,and
29、 they seem willing to do the hard work that needs to be done.Foundation model improvementsincluding domain and industry customizationand the promise of AI agents could help overcome inherent challenges and accelerate the creation of business value.However,it might be a multiyear journey for some org
30、anizations to reach full-scale deployment and achieve the ROI they are looking for.With GenAI,some level of uncertainty is unavoidable and the technology will likely continue to advance at a rapid pace.Business and technology leaders,for their part,should focus on what they can controlnamely,organiz
31、ational readiness,particularly in areas such as data,risk management,governance,regulatory compliance and workforce/talent.Addressing issues in these key areas will help position organizations for success with GenAI no matter how the future unfolds.IntroductionKey findingsAbout the State of Generati
32、ve AI in the Enterprise:Wave four survey resultsThe wave four survey covered in this report was fielded to 2,773 director-to C-suite-level respondents across six industries and 14 countries between July and September 2024.Industries included:consumer;energy,resources and industrials;financial servic
33、es;life sciences and health care;technology,media and telecom;and government and public services.The survey data was augmented by additional insights from 15 interviews with C-suite executives and AI and data science leaders at large organizations across a range of industries.For details on methodol
34、ogy,please see p.45.This quarterly report is part of an ongoing series by the Deloitte AI InstituteTM to help leaders in business,technology and the public sector track the rapid pace of Generative AI change and adoption.The series is based on Deloittes State of AI in the Enterprise reports,which ha
35、ve been released annually the past five years.Learn more at case studies featured in this report are a small subset of the insights from our ongoing in-depth interviews with business and AI leaders from a wide range of industries.The goal is to build on the quantitative findings from our quarterly s
36、urveys by capturing practical,real-world insights directly from leaders and organizations on the front lines of GenAI adoption.Our interviews explore how leading organizations in diverse industries are using GenAI to create value.Most notably,we are seeing initiatives focused on applying GenAI to bu
37、siness-specific challenges in areas critical to success in that organizations industry.Examples include using GenAI for:Brand promotion and integrated business planning in the consumer products industry Predictive maintenance for physical assets in the energy industry Drug discovery and clinical tri
38、al tracking in the pharmaceutical industry Cybersecurity and portfolio management in the financial services industry Sales enablement,chip development and improved search in the technology industry Archive management and music source separation in the media and entertainment industryThis focus on mi
39、ssion-critical activities suggests a broad strategic shift in the GenAI landscape,from technology catch-up to competitive differentiation.Real-world case studiesGo to case studies8 Looking back at 20249 9 9Our first global quarterly survey,conducted in late 2023,revealed great excitement and expecta
40、tions for GenAI.However,those feelings were tempered by uncertainty and fear about the technologys potentially negative impacts on workers and society.Our second and third quarterly surveys focused more deeply on how organizations were prioritizing tangible results and value creation from their GenA
41、I investments,and on understanding and tackling the barriers to successful scaling.A key finding during the year was that promising results from early GenAI pilots were raising expectations and driving increased investment in the technology.Today,interest and excitement about GenAI remain high.Howev
42、er,the initial fervor has gradually given way to a positive yet pragmatic mindsetespecially among business leaders at all levels.Meanwhile,technology leaders interest and excitement have remained high and steady(figure 1).Although this shift among business leaders might seem like a step backward for
43、 GenAI,it is entirely consistent with the usual life cycle for transformative technologies.It is also a net positive in terms of helping organizations move past the hype stage so they can directly tackle the serious work of using GenAI to create real business value.Looking back at 2024Now:Looking ba
44、ck at 2024Figure 1Q:For the following groups in your organization,rate their overall level of interest in Generative AI.State of Generative AI in the Enterprise Survey,Q1(Oct./Dec.2023)N(Total)=2,774;Q4(July/Sept.2024)N(Total)=2,773;14 countries common to both data setsLevel of interest in GenAI(hig
45、h+very high)Q1Q4BoardC-suite/executive leadersTechnical leadersLOB/functional leadersEmployees62%74%86%64%49%46%59%86%56%50%A key finding during the year was that promising results from early GenAI pilots were raising expectations and driving increased investment in the technology.-16 pts-15 pts1010
46、Over the past year,as organizations gained experience with GenAI,they began to better understand both the rewards and challenges of deploying the technology at scaleand adjusted their plans and expectations accordingly.Budgets have risen,and the need for C-suites and boards to spur their organizatio
47、ns into action has diminished.At the same time,the need for disciplined action has grown.Technical preparedness has improved,while regulatory uncertainty and risk management have become bigger barriers to progress.Talent and workforce issues remain important;however,access to specialized technical t
48、alent no longer seems to be the dire emergency it once was,at least in comparison to other priorities.There has been one constant,however:improved data management continues to be a top priority,even for companies that live and breathe data.“Data emerged as the central factor for our GenAI success,”s
49、aid a former software engineering manager for one of the worlds leading technology companies.“While the models and computing power existed,accessing the right data proved to be the biggest bottleneck.To address this,the company implemented a centralized data strategy,managed by a single data leader,
50、to streamline data acquisition and minimize redundancyenabling faster model development.”Now:Looking back at 2024“Data emerged as the central factor for our GenAI success”Former software engineering manager for leading technology company11From a technology perspective,the capabilities of foundation
51、models and applications have improved dramatically over the past year.There are smaller,more efficient models;better latency;bigger access windows;expanded modalities;greater autonomy;and increased model specialization.Reliability and trust have improved as well,although both still have a long way t
52、o go.Meanwhile,the adoption rate for customized,open-source and/or proprietary large language models(LLMs)remains limited at 20%25%of those surveyed.Over the past year,respondents reported they believe their organizations have most improved their GenAI preparedness in the critical areas of technolog
53、y infrastructure(+7 points)and strategy(+5 points).However,preparedness has seemingly not improved in the other critical areas of risk and governance and talent.The vast majority of respondents(78%)reported they expect to increase their overall AI spending in the next fiscal year,with GenAI mostly e
54、xpanding its share of the overall AI budget relative to our first-quarter survey results.In particular,the percentage of organizations investing 20%39%of their overall AI budget on GenAI climbed by 12 points,while the percentage of organizations investing less than 20%of their AI budget on GenAI fel
55、l by 6 points.“The way we do business has not changed,”said the VP of artificial intelligence at a major media and entertainment company.“For every project,our objective is always to do something that has a positive impact on the business.This has not changed and is not going to change because its w
56、hat makes sense.However,a large proportion of project proposals now have a GenAI component to them.”Now:Looking back at 202478%of respondents expect to increase their overall AI spending in the next fiscal year.12Relative to other respondents,the C-suite leaders(CxOs)in our survey generally demonstr
57、ated higher levels of excitement and optimism about their organizations GenAI implementations.For example,21%of C-suite survey respondents reported they feel GenAI is already transforming their organization,compared to only 8%of non-C-suite respondents.C-suite executives surveyed are comparatively l
58、ess worried about barriers such as trust,risk management,governance and regulatory compliance.They also have a rosier view of how quickly their organization is moving,and how quickly the barriers to scaling and value creation will be addressed.Sixty percent of non-C-suite respondents believe it will
59、 take 12 months or more to overcome scaling barriers,compared to only 47%of C-suite respondents.This doesnt necessarily mean CxOs are out of touch with the challenges of adopting and deploying GenAI.It could be they are still playing the primary role of catalyst or cheerleader and are in the process
60、 of learning what it really takes to implement and scale GenAI.What will be important going forward is for CxOs to direct that enthusiasm to removing barriers and enabling scaling.Now that GenAI in the enterprise is moving past its infancy,CxOs should take on new roles,including those of guide,couns
61、elor and challenger.Chief executive officers should show top-down support for GenAI,be the champions for governance and risk initiatives,and foster an environment of trust and transparency.Chief information officers,chief technology officers and chief data officers should sharpen their focus on iden
62、tifying and overcoming the barriers to large-scale GenAI deployment within their domains.Chief financial officers should ensure responsible spending without stifling innovation.And chief human resource officers should promote training,reskilling and other human capital investments.View from the C-su
63、iteNow:Looking back at 202413The uneven pace of changeWith transformational technologies,there are always gaps between the pace of technological change and the ability of individuals,businesses and policymakers to keep up.GenAI is no exception.Incredible advances in GenAI technology,fueled by massiv
64、e capital and intellectual investments from tech companies,are already manifesting in individuals everyday livesthrough smarter smartphones,improved customer service,AI-enhanced search engines,and more.For businesses,embracing and integrating GenAI is much harderand takes much longerdue to a complex
65、 mix of factors.This could include dealing with competing transformational priorities.However,policy,legislative and regulatory changes might be more challenging overall.Governments today face the monumental task of regulating a technology whose capabilities are still taking shape.One direct consequ
66、ence is that regulatory compliance has emerged from the pack to become the top barrier holding organizations back from developing and deploying GenAI tools and applications(figure 2).This highlights respondents unease about which use cases will be acceptable,and to what extent their organizations wi
67、ll be held accountable for GenAI-related problems.This uneven pace of change creates friction for organizations,which likely contributes to the relatively moderate pace of transformation we are seeing as businesses work through their challenges on the path to creating sustained value with GenAI.Now:
68、Looking back at 2024Barriers to developing and deploying GenAIQ:What,if anything,has most held your organization back in developing and deploying Generative AI tools/applications?(Select up to three challenges)State of Generative AI in the Enterprise Survey,Q1(Oct./Dec.2023)N(Total)=2,774;Q4(July/Se
69、pt.2024)N(Total)=2,773;14 countries common to both data setsFigure 2Worries about complying with regulationsDifficulty managing risksLack of an adoption strategyDifficulty identifying use casesTrouble choosing the right technologiesImplementation challengesLack of technical talent and skillsLack of
70、a governance modelCultural resistance from employees28%Not having the right comp.infrastructure/dataLack of executive commitment and/or funding38%26%32%26%27%36%26%27%24%18%22%25%21%20%17%17%17%15%15%19%14%Q1Q4+10 pts.+6 pts.-10 pts.14 Now:Where we are1515For our fourth wave report,we wanted to answ
71、er several questions about scaling and value realization.Where do things stand with workforce adoption?How many experiments are organizations pursuing,and what are their success rates?Which benefits are GenAI initiatives targeting?Are some types of GenAI initiatives/use cases showing more promise th
72、an others?Are they meeting ROI expectations?Now:Where we are1234516Where do things stand with workforce adoption?Now:Where we are1Our latest survey results show that access to GenAI is still largely limited to less than 40%of the workforce.Also,for most organizations,fewer than 60%of workers who hav
73、e access to GenAI actually use it on a daily basis.This suggests many companies have yet to integrate GenAI into their standard business workflows.It also raises the chicken-and-egg question of whether limited access to GenAI is inhibiting comfort and uptake with the technology(and stifling innovati
74、on),or whether the lack of high-value,innovative use cases is limiting interest and adoption.For GenAI to become truly transformational,it will likely require greater numbers of workers experimenting and leveraging the technology to identify new,high-impact use cases within the business.“Within our
75、organization,the demand for GenAI use cases and innovation primarily comes from middle management and employees,rather than being driven by the C-suite,”said the director of product management for GenAI,cloud and data centers at a leading semiconductor company.“While the C-suite has been slower to e
76、ngage in AI implementation,teams across the company are developing proofs-of-concept and driving AI adoption through internal boards and governance structures.This bottom-up approach emphasizes improving workflows and test cases,with leadership providing support as needed for broader integration.”Of
77、 course,access alone does not equate success.Providing access to GenAI does not mean workers will use it.Conversely,workers with a burning desire to use GenAI will likely find a way to do so,with or without approval.However,in order to foster transformation and maintain some level of control over ho
78、w GenAI is used within the enterprise,it generally makes sense to offer broad workforce access to sanctioned GenAI tools,supported by clear guidelines for proper use.“Currently,GenAI adoption is driven by internal demand,with early adopters seeking to use the tools to meet their specific needs,”said
79、 the head of GenAI in product management at a major technology company.“However,we expect a shift towards push-driven adoption in the next year,where all business units will be required to integrate the platform as it becomes an approved and proven tool.This shift will create pressure for teams to l
80、everage the technology or risk missing out on the benefits it offers.”“Currently,GenAI adoption is driven by internal demand,with early adopters seeking to use the tools to meet their specific needs”Head of Generative AI,project management at major technology company17We found organizations are stil
81、l heavily experimenting with GenAI,and scaling tends to be a longer-term goal.Over two-thirds of respondents said that 30%or fewer of their current experiments will be fully scaled in the next three to six months.This suggests companies are taking time to test GenAIs capabilities and to figure out w
82、here it can help the most(figure 3).The lions share of organizations are currently pursuing 20 or fewer GenAI experiments or proofs of concept(POCs)and expect to fully scale 10%30%of those experiments in the next three to six months.As expected,individual company actions vary,with larger numbers of
83、experiments being conducted by organizations that are large,advanced in their use of AI,and/or operating in key industries of technology,media and telecommunications;life sciences and health care;or financial services.What is the state of GenAI experimentation?Q:Approximately how many Generative AI
84、experiments or proofs of concept is your organization currently pursuing?What percentage of these AI experiments or proofs of concept do you anticipate will be fully scaled in the next three to six months?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,773Figure 3Now:Where
85、 we are2Volume of experiments/POCs3%More than 10051 to 10021 to 5011 to 20Less than 10Dont know7%35%24%29%3%Volume of experiments/POCs%of organizationsScaling progress(next 3-6 months)2%80%2%9%5%13%26%of experiments/POCs27%16%1%of organizations70%60%50%40%30%20%10%0%18“Improved efficiency and produc
86、tivity”continue to be the most commonly sought benefits from GenAI,and many organizations(40%)reported they are already achieving their expected benefits in this area to a large or very large extent.However,our respondents cited slightly higher levels of success in a small handful of more strategic
87、benefit areas,particularly“new ideas and insights”(46%)and“innovation and growth”(45%)(figure 4).Which benefits are GenAI initiatives targeting?Now:Where we are60%50%40%30%20%10%10%15%20%25%30%35%40%45%50%55%60%Benefits achieved (among companies that sought it,the%that achieved it to a large or very
88、 large extent)Benefit sought(%hoping to achieve the benefit)Q:What are the key benefits you hope to achieve through your Generative AI efforts?(Select up to three benefits)To what extent are you achieving those benefits to date?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)
89、=2,773Figure 4Benefits achieved vs.benefits soughtDetect fraud and manage riskachievingseekingIncrease speed/ease of developing new systemsEnhance relationships with clients/customersUncover new ideas and insightsEncourage innovation and growthImprove efficiency and productivityImprove existing prod
90、ucts and servicesShift workers from lower-to higher-value tasksIncrease revenueReduce costs346%of respondents(seeking the benefit)reported that they are uncovering new ideas and insights with GenAI.19To understand where GenAI is having the deepest impact on organizations,we asked respondents to cons
91、ider one of their most advanced GenAI initiativesan initiative that is most fully scaledand then to identify which function or department it targets.Since GenAI is a highly advanced technologyand one of its best capabilities is generating computer codeits no surprise that the IT function came out on
92、 top(28%).However,the survey data also shows GenAI being deployed deeply in many other parts of the business as well,including operations(11%),marketing(10%),and customer service(8%)(figure 5a).Figure 5aQ:Consider one of your organizations most advanced(scaled)GenAI initiatives.In which function or
93、department is this initiative?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,773GenAI initiatives are most advanced within these functions 28%ITOperationsMarketingCustomer serviceCybersecurityProduct developmentAre some use cases showing more promise?Now:Where we areR&DSa
94、lesStrategySupply chainFinanceHRManufacturingLegal,risk,compliance 11%8%10%8%6%7%5%4%5%4%2%2%1%420Even more revealing,we found that the most advanced GenAI applications outside of IT overwhelmingly target critical business areas that are fundamental to success in a companys specific industry(e.g.,ma
95、rketing in the consumer industry;operations in energy,resources and industrial;cybersecurity in financial services).For example,in the life sciences and health care industry,where R&D is strategically important,the associate director of artificial intelligence at a leading health care products compa
96、ny said:“Value creation is measured operationally by the acceleration of development timelines,with AI providing faster results while staying within set performance and output quality constraints.Our focus is on development speed,rather than outperforming human capabilities.And while a tenfold accel
97、eration without human involvement remains aspirational,a three-to five-fold increase in speed has already been realized.”This is a crucial insight since many business leaders still associate GenAI with personal productivity and other relatively mundane tasks secondary to the core business.“Our compa
98、ny has an enterprisewide AI leadership team,but I think theyre really focused on a co-pilot strategy and helping all individuals use AI tools to improve their productivity,”said the director of organizational transformation and change at a leading consumer products company.“Were a little bit behind
99、the eight ball on internal processes,and AI is sort of on the fringe.I dont think business-facing case studies have been weaved into an overall enterprise AI strategy.”Now:Where we areTop three most advanced(scaled)GenAI initiatives by industryColor of the bubble represents the functionTech,media&te
100、lecomGovernmentIT 96%Operations 3%IT 34%Product dev 17%Cybersecurity 12%IT 23%R&D 21%Operations 11%IT 21%Cybersecurity 14%Finance 13%Operations 23%IT 17%Strategy 11%IT 20%Marketing 20%Customer service 12%Figure 5bQ:Consider one of your organizations most advanced(scaled)GenAI initiatives.In which fu
101、nction or department is this initiative?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,773IndustryTop 3 functions using GenAI applications and the percentage of initiatives in eachConsumerEnergy,resources&industrialFinancial servicesLife sciences&health care21Are advanced
102、 GenAI initiatives meeting ROI expectations?Return on investment for organizations most advanced GenAI initiatives has been generally positive.Almost all organizations report measurable ROI,and one-fifth(20%)report ROI in excess of 30%.Similarly,nearly three-quarters(74%)say their most advanced init
103、iative is meeting or exceeding their ROI expectations(43%meeting,31%exceeding).Also,two-thirds(67%)say their most advanced initiative is at least moderately integrated into their broader work processes(figure 6).Figure 6Most advanced(scaled)GenAI initiativesNow:Where we are51%or more 31%to 50%11%to
104、30%6%to 10%Less than 5%Not measuring ROI to date 6%14%23%41%9%5%Significantly aboveSomewhat aboveMeetingSomewhat belowSignificantly belowROI expectations 7%24%19%43%5%Completely integratedLarge extentModerate extentSmall extentNot at all,but intend toNo intention to integrateLevel of integration 4%2
105、0%25%43%7%2%Q:ROI to date:Estimate the ROI to date for this specific initiative./ROI expectations:How is the ROI from this Generative AI initiative meeting your organizations expectations?/Level of integration:To what level is the Generative AI initiative integrated into the broader organizations wo
106、rk process?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,77374%of respondents say their most advanced Generative AI initiative is meeting or exceeding their ROI expectations.522Relative to other types of advanced GenAI initiatives,those focused on cybersecurity are far m
107、ore likely to be exceeding their ROI expectations,with 44%of cybersecurity initiatives delivering an ROI somewhat or significantly above expectations versus only 17%that are delivering an ROI somewhat or significantly below expectations(a 27-point gap)(figure 7).On the other hand,with advanced GenAI
108、 implementations in functions such as sales,finance and R&D,more respondents reported ROI below expectations than reported ROI above expectations.This suggests some challenges have yet to be overcome in those areas.Meanwhile,36%of respondents said their cybersecurity initiative is integrated into wo
109、rk processes to a large extenta higher level of integration than any other kind of advanced GenAI initiative.These results are somewhat skewed by advanced GenAI deployments in the financial services and technology industries,where cybersecurity is especially critical.However,the relatively strong pe
110、rformance of cyber-related GenAI initiatives makes sense for a number of other reasons as well.1 Many organizations are already experienced in using AI to manage cyberthreats and have related infrastructure in place to scale cyber capabilities.According to Deloittes Global Future of Cyber Survey,fou
111、rth edition,86%of the organizations surveyed already deploy AI-based tools to continuously monitor their digital infrastructure to a moderate or large extent.2Now:Where we are45%40%35%30%25%20%15%10%20%25%30%35%40%45%50%Below ROI expectations(%saying somewhat and significantly below)Above ROI expect
112、ations(%saying somewhat and significantly above)Q:How is the ROI from this Generative AI initiative meeting your organizations expectations?State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,773Figure 7ROI performance against expectations (for most advanced initiatives)R&D-5O
113、perations 0Sales-8Finance-8More below than aboveMore above than belowProduct dev 0Marketing+2Strategy+10Customer service+8Supply chain+7IT+15Cybersecurity +27Those focused on cybersecurity are far more likely to be exceeding their ROI expectations.Meeting expectationsSignificantly above expectations
114、Below expectationsAbove expectationsThe numbers with each function indicate the difference between the%above expectations and the%below expectations.23 Next:Looking ahead2424In our final survey of the year,we wanted to explore how organizations expect GenAI adoption and value creation to unfold over
115、 the next 1218 months.What do they think could slow adoption?How long do they expect it will take to overcome their GenAI-related challenges,and are they willing to sustain their commitment long enough for their investments to pay off?Also,what emerging GenAI technologies are they most interested in
116、?What could slow GenAI adoption?According to our respondents,there are a range of issues with the greatest potential to slow overall marketplace adoption of GenAI over the next two years.The top potential barriers to adoption include:mistakes/errors with real-world consequences(35%);not achieving ex
117、pected value(34%);shortage of high-quality data(30%);and general loss of trust due to bias,hallucinations and inaccuracies(29%)(figure 8).For broader GenAI adoption to occur,the technologys reliability,accuracy and trustworthiness will need to improve.Also,GenAI initiatives will need to deliver thei
118、r expected value in a timely manner.Next:Looking aheadQ:Which of the following do you think could MOST slow adoption of GenerativeAI models/tools/applications by organizations over the next two years?(Select two)State of Generative AI in the Enterprise Survey,(July/Sept.2024)N(Total)=2,773Figure 8Im
119、pediments to GenAI adoption in the near future35%Mistakes/errors leading to real-world consequencesNot achieving expected valueThe availability of enough high-quality dataA general loss of trust due to bias,hallucinations and inaccuraciesIntellectual property issues34%30%29%25%Concerns over energy u
120、sage/environmental concernsRising cost of building and operating models/tools/applications20%13%AI sovereignty issues12%35%of organizations we surveyed said their top potential barrier to adopting Generative AI is mistakes/errors with real-world consequences.25As noted above,“not achieving expected
121、value”is in a virtual tie as the No.1 potential barrier to overall adoption of GenAI.Yet,the majority of respondents(55%70%,depending on the challenge)believe their organizations will need at least 12 months to resolve adoption challenges such as governance,training,talent,building trust and address
122、ing data issues(figure 9).According to the head of finance for private assets investments and strategic ventures at a leading financial services company,“To create value from our GenAI use case,we will need to fundamentally transform our operating cost model by reducing fees and demonstrating that o
123、ne portfolio manager can manage multiple portfolios efficiently over time.This will take at least five years to validate and substantiate the KPIs fully.”Will organizations have the patience and sustained commitment to work through their GenAI challenges,or will they cut and run before their investm
124、ents have a chance to pay off?In our latest survey,70%of respondents said their organizations will need at least 12 months to resolve the challenges related to surpassing or achieving their expected ROI from GenAI.However,76%reported their organizations will wait at least 12 months before reducing i
125、nvestment in a GenAI initiative that is not meeting its value targets.Based on these two responses,it appears organizations will likely have the patience necessary to see their GenAI investments pay off.How long will it take to resolve challenges related to GenAI,and are organizations willing to wai
126、t?Time to resolve GenAI challengesLess than 1 year1 to 2 yearsFigure 9Q:With respect to your organizations priority Generative AI initiatives,when do you think the organization will adequately resolve challenges around the following areas?State of Generative AI in the Enterprise Survey,(July/Sept.20
127、24)N(Total)=2,773More than 2 yearsImplementing a governance strategy29%52%17%Achieving ROI26%55%15%Acquiring talent37%51%9%39%49%9%Addressing data challenges40%51%Overcoming scaling barriers42%48%Training workers44%48%Next:Looking ahead8%8%7%Managing trust issues26Among all the emerging GenAI-relate
128、d technological innovations,agentic AI currently appears to be capturing the most interest and attention.In fact,according to our survey,the two most interesting areas today are agentic AI(52%)and multiagent systems(45%),which is essentially a more advanced,complex variant of agentic AI.Closely behi
129、nd those two is multimodal capabilities,which is also an integral part of agentic AI systems(figure 10).Interest in future GenAI-related developmentsFigure 10Q:What Generative AI technology developments is your organization most interested in?(Select all)State of Generative AI in the Enterprise Surv
130、ey,(July/Sept.2024)N(Total)=2,77352%GenAI for automation(agentic AI)45%Multiagent systemsSynthetic data for training/tuningLarge action modelsAdvanced hardware specifically for GenAI applications44%Multimodal capabilitiesNew training techniques35%Smaller,less resource-intensive models35%30%Alternati
131、ve/improved architectures28%21%Next:Looking aheadWhich technology advances could drive the future of GenAI?20%27AI agents are software systems that can complete complex tasks and meet objectives with little or no human intervention.They are called“agents”because they have the agency to act independe
132、ntly,planning and executing actions to achieve a specified goal.3 The vision for agentic AI is that autonomous AI agents will be able to execute assigned tasks consistently and reliably by acquiring and processing multimodal data,using various tools to complete tasks,and coordinating with other AI a
133、gentsall while remembering what theyve done in the past and learning from their experience.“In the next phase of GenAI,we envision the development of specialized AI agents tailored to specific functions,like sales research,to manage the overwhelming volume of data,”said the director of product manag
134、ement for GenAI,cloud and data centers at a leading high-tech manufacturing company.“These agents will streamline processes,helping sales teams gather critical information quicklywithout the need for extensive manual research.Multiagent workflows are a future possibility;however,we anticipate starti
135、ng with single-agent solutions that can mature and scale efficiently,focusing on ROI as they evolve into production.”Agentic AI is the next logical step for GenAI,giving GenAI-based systems access to more types of information and increasing AIs level of responsibility and autonomy.In fact,26%of our
136、survey respondents said their organizations were already exploring autonomous agent development to a large or very large extent.However,as with current GenAI systems,agentic AI is not a silver bullet for everything a company needs to get done.The key barriers currently facing GenAIsuch as regulatory
137、 uncertainty,inadequate risk management,data deficiencies,and workforce/talent issuesstill apply and are arguably even more important and challenging due to the increased complexity of agentic AI systems.Next:Looking ahead“In the next phase of GenAI,we envision the development of specialized AI agen
138、ts tailored to specific functions,like sales research,to manage the overwhelming volume of data.”Director of product management for GenAI,cloud and data centers at a leading high-tech manufacturing company28 Considerations2929Initially,senior executives acted as catalysts and drivers for GenAI adopt
139、ion in their organizations.However,with strategies set,funding approved and guidance given,many are now expecting GenAI to deliver significant and timely improvements in efficiency,productivity,innovation and competitive advantage.As such,C-suite leaders(CxOs)today should think about how to redefine
140、 their roles around GenAIand how to best lead their organizations forward.There are three main ways CxOs can aid in this preparation.First,they must ensure the organization stays aligned.Technical and business executives should be involved in each others conversations and decisions,making sure GenAI
141、 is appropriately represented.Second,CxOs must manage organizational expectations.Leaders at the most senior level tend to be more optimistic than those below them when it comes to the organizations rate of progress with GenAI(and ability to overcome obstacles).The GenAI journey is long,and C-suite
142、leaders need to be realistic about time horizons for project success and organizational transformation.Third,CxOs must show patience in the face of uncertaintyproviding a steady hand and sustained commitment to achieving long-term transformation across multiple business areas.Task the C-suite with c
143、reating alignment and managing expectations Next:Considerations30GenAI initiatives are already delivering significant enterprise value,including improved efficiency,relationships and innovation.However,our survey results show that measurable ROI varies widely for different use cases and functions.So
144、me initiatives are already exceeding expectations,but others are currently falling short.The bridge to sustained ROI can only be built by establishing the right holistic strategies,building platform capabilities,being realistic about targets and timelines,and taking some risks.In our case studies,we
145、 found that focusing on a small number of high-impact use cases in proven areas can accelerate ROI,as can layering GenAI on top of existing processes.Additionally,centralized governance can pave the way for smoother adoption and employee buy-in,which tends to yield better results and improves scalab
146、ility.Finally,continuous iteration based on user feedback and real-world performance can help ensure sustained value creation.Ultimately,organizations need to move beyond isolated initiatives and integrate GenAI into increasingly sophisticated and interconnected processes,evolving toward cognitive s
147、ystems with advanced reasoning capabilities.The goal should be to fundamentally reinvent business processes.Build bridges to sustained ROIPrioritize your workforce and prepare it for disruptionAccording to our survey results,the number of organizations that feel prepared for GenAI from a talent pers
148、pective is still quite low and hasnt changed much since the beginning of 2024.Also,workforce access to GenAI tools is still somewhat limited and daily use remains low.These results all shine a spotlight on the need for organizations to do more to prepare their workers for potential disruption from G
149、enAI.Although organizations have many priorities and barriers to focus on,they cant overlook talent issues if they want to achieve sustained growth and maximize ROI.Workers need more GenAI access and experienceand they need it sooner rather than later.Several of our case studies revealed organizatio
150、nal resistance to adopting GenAI solutions,which slowed project timelines.Usually,the resistance stemmed from unfamiliarity with the technologies and/or skill and technical gaps.Effective change management,including education and training,was pivotal in overcoming the challenge.Without adequate work
151、force buy-in and training,even the most powerful GenAI solutions can fail to deliver the expected outcomes.Also,developing systems for continuous improvement is criticalwith users providing ongoing feedback on the quality and accuracy of GenAI solution outputs.Next:Considerations31With agentic AI,th
152、e question is not if,but when.Although the technology is still in its early stages,it is evolving rapidly and will likely become increasingly capable over the next few years.And while there are still many challenges to overcomeand technical complexities to sort outnow is the time to start preparing.
153、4 Organizational knowledge and experience gained from GenAI implementations will help with the development and deployment of AI agents.Also,the 13 elements of scaling mentioned in our prior GenAI reports will be just as applicable to agentic AI.5 Organizations can begin by developing a strategic roa
154、d map and assessing which tasks and workflows are well-suited for agentic AI.Identify specific goals and desired value.Map out the risks associated with autonomous agents and create mitigation plans.Start with low-risk use cases that use noncritical datawith human oversight as a backup.These early s
155、teps can help test and build the data management,cybersecurity and governance capabilities necessary for safe agentic AI applications.Once your organization is comfortable,it can then progress to applications that use more proprietary data,have access to more tools,and operate more autonomously.Star
156、t planning for GenAI agents Manage an uncertain futureGenAIs present is filled with great promise,but its future holds many uncertainties.Will investments pay off in the long term?Will bias,hallucinations,misinformation and“AI-generated pollution”be controlled?Will GenAI use cases lead to new busine
157、ss models and breakthrough innovations or just optimize existing operations?How fast will GenAI achieve broad,human-level performanceif ever?Although no one can answer these questions,one thing we know for sure is that all the uncertainty surrounding GenAI is hindering its progress.To act confidentl
158、y and decisively in the face of this uncertainty,organizations should consider boosting their efforts and capabilities in the areas of foresight,market sensing and scenario planning.6 This will help leaders model plausible futures,identify potential blind spots in their strategies,and make more info
159、rmed decisions today.The widespread transformation being driven by GenAI is truly an odyssey that will take place over many years and have many phases.Building the right capabilities today will help your organization make more informed strategic choices and position itself to capitalize on future de
160、velopments and opportunities.Next:Considerations32 Case studies3333GenAI is boosting software security in banking In banking,robust cybersecurity and data governance are essential to protect sensitive customer data,comply with complex regulatory frameworks,and maintain public trust.Case study 1:Retu
161、rn to page 834We met with the global head of GenAI,cloud and data privacy at a leading bank to explore how GenAI is transforming secure software development in financial services.By analyzing application vulnerability alerts and reducing false positives,GenAI enables engineers to focus on critical i
162、ssues,limit the number of actionable alerts and enhance operational efficiency.Problem On a typical day,the banks security team faces millions of alerts related to code-level security issues,such as endpoint vulnerabilities and misconfigurations.Managing this volume of alerts is both time intensive
163、and yields false positives,leading to tension with the application developers whose performance incentives are aligned with new feature development rather than vulnerability remediation.“Previously,developers got frustrated because 80%of their time was spent remediating vulnerabilities.Their perform
164、ance is measured by how many new features they deliver,not how many vulnerabilities they fix in their code,”said the leader we interviewed.Solution The banks solution aimed to improve the way software is securely developed with GenAI.The leader explained that the solution was built on a mature AI fo
165、undation within the bank.The team deployed“an AI-powered platform,which translates regulations,policies and standards into security controls(including preventative controls,detective controls,responsive controls and corrective controls),and then codifies those controls across the software developmen
166、t life cycle.”From there,facing a daily deluge of potential application security alerts,the bank needed an efficient yet accurate way to identify critical vulnerabilities.To address this need,the banks security operations center implemented a GenAI solution to streamline its vulnerability management
167、 processes and systems.Case study 1:GenAI is boosting software security in banking35Approach The solution triages millions of incoming cyberthreat alerts,paring them down to thousands of“real threats”that then go to different cyber teamsfor example,distributed denial-of-service,malware and others.To
168、 enable that prioritization,different security control requirements are assessed to score and reduce those alerts down to the most critical threats based on breachability(the size of the risk)and exploitability(the likelihood of exploitation by a threat actor).Additionally,as GenAI is increasingly u
169、sed to translate regulatory requirements,controls can become more automated.For example,GenAI can summarize requirements such as the need to rotate encryption keys at set intervals and identify opportunities to automate the banks security protocols,or it can be used as an intelligence-gathering tool
170、 to identify common security risks that should be automated.For example,“Say an employees login credentials arent used for more than 30 days;AI can detect that and disable the account,”said the leader.“This reduces cybersecurity risk by reducing the attack surface.”Results When asked how to think ab
171、out ROI for this type of solution,the bank leader explained,“We calculate the cost for the potential risk against the cost of remediating this risk.”For security,the risk economic model covers domains positively impacted/measured by the banks data-driven,risk-based,decision-making process.These doma
172、ins include data protection,encryption,address in transit,in use,network segmentation,authentication,authorization,logging and monitoring.The solution has dramatically reduced the number of common application security vulnerability alerts the cyber team must triage and development teams must address
173、down to fewer than 10 critical vulnerabilities a day.Overall,the GenAI solution has significantly reduced the banks cyber risk by enabling its security and development teams to focus their time and effort on problems that are real,impactful and actionable.It has also boosted morale and productivity
174、across the engineering team by reducing the time spent on DevSecOps so they can focus more time on what theyre economically incentivized to dodevelop new software and push critical updates into production.Case study 1:GenAI is boosting software security in banking36GenAI is accelerating sales succes
175、s in tech Tech companies are players on both sides of the Generative AI market,developing GenAI-based products and services they can sell to external customers while also harnessing the power of GenAI to help their own workers and enhance their own business processes.Case study 2:Return to page 837W
176、e spoke with the head of Generative AI product management at a large tech company to learn how his group is using GenAI.He described how the company uses a centralized process to collect all internal GenAI use cases from various business units,and then prioritizes them based on importance and feasib
177、ility.Use cases are categorized into three types:(1)external-facing tools such as chatbots and“agentic solutions”aimed at improving customer service;(2)internal developer tools or“co-pilots”designed to enhance productivity;and(3)playgrounds driven by application programming interfaces(APIs)that allo
178、w developersincluding technical and business usersto build custom applications for specific needs not covered by the other two categories.By employing a structured,centralized approach,the company aligns GenAI projects with core business goals,ensuring high relevance and strategic impact across mult
179、iple functions.One compelling example from the API playground category is the firms accelerated sales application,enabled by GenAI.The solution aims to make the companys sales teams more efficient and effectiveand help them close deals fasterwith an eye toward eventually selling those same capabilit
180、ies as an external product.Problem When it comes to selling big tech,time is money.Sales reps need to use their time wisely so they can pursue more deals and build stronger relationships with clients.Although they have access to detailed playbooks and other materials designed to help them sell more
181、effectively,sales reps struggle with inconsistent processes and dispersed resources,making it challenging to efficiently access the available information.Whats more,sales and marketing leaders have different intake points across different business units,which makes for a highly variable process.Sale
182、s reps also must be very timely when responding to a new opportunity,especially a“tight deadline”request for proposal(RFP).“External customers often need to spend their budgets quickly,otherwise the budget will be gone,”said the executive we interviewed.“In many cases,the window to respond to an RFP
183、 is just three or four business days.”Solution The companys new GenAI-powered sales tools have two major components.One is an RFP response tool that allows sales reps to summarize customer requirements and expedite the creation of responses to RFPs,allowing business leaders to more quickly generate
184、a complete and customized proposal with just a few mouse clicks.The other is an interactive chatbot with access to the companys internal knowledge base of playbooks and other sales materials.The solution helps business leaders quickly summarize information to better prepare for pitches.“Using the to
185、ol is very similar to other chatbot experiences,but its more within our internal domain,”the executive said.“Imagine Im a sales rep and tell the chatbot,I want to sell X,Y,Z.What is the playbook?And the system responds by giving me some customized bullet points I can rehearse with myself before I ha
186、ve to pitch to the client.”Case study 2:GenAI is accelerating sales success in tech38Approach The overall strategy originated from the CEOs dual GenAI agenda of improving internal productivity and identifying external commercialization opportunities.This approach included guidance to develop an inte
187、rnal platform that,if proven,could potentially be offered in the marketplace to external clients facing similar challenges.The company used a“sandbox”approach when developing the new sales tools.This gave interested sales reps access to GenAI tools and APIs in a safe,low-cost environment so they cou
188、ld experiment freely and develop new applications without writing computer code.The solution aims to detect common customer pain points and then use those insights to generate sales activity by identifying opportunities for commercialization or optimization.Results The GenAI tools seek to enhance de
189、al closure speed and size and improve the accuracy of proposal generationwith the ultimate goal of increasing sales performance by leveraging internal knowledge resources more effectively.Currently,the company is more focused on feasibility than monetary returns,using key performance indicators(KPIs
190、)related to efficiency and time savings.How many sales reps are using the tool,and how long is the period between their first access and generating their first output?If the period is short,its a sign the tool is both appealing and easy to use.The earliest measure of success is onboarding.Of course,
191、the most important measure of success for a sales tool is its impact on sales.In conducting direct A/B comparisons between sales processes that used the new GenAI-powered tools and those that didnt,the company found a marked improvement in how quickly deals got closed when the tools were used.Althou
192、gh the tools are not yet ready to be offered as external products,doing so remains a top priority.According to the executive we interviewed,“Our companys CEO would really like us to first adopt this Generative AI platform internally and then try to think about any way we can sell it to external cust
193、omers.”What does long-term success look like for its GenAI sales tool?The hope is for increased deal sizes,faster deal closings,and effective deployment of a commercialized external solution.Case study 2:GenAI is accelerating sales success in tech39GenAI is powering an always-on,multimodal social me
194、dia presence in the consumer industry Social media is an increasingly important marketing channel for all consumer companies,allowing them to convey the voice of their brand and reach customers in a highly compelling way.Case study 3:Return to page 840We spoke with the senior director,head of data a
195、nd analytics for a leading global consumer company to learn how his team is activating a GenAI strategy to help the companys brands fully automate and expand the scope of their real-time social media trend analysis and content creation.Problem Social media marketing is a critical business activity t
196、hat is costly,time-consuming and subject to human bias.In a recent year,social media strategy and content generation cost the company US$500M,with much of that spent on third-party contracts with media and creative agencies.Solution The company is now using GenAI to produce and manage much of its br
197、and-focused social media content,including copywriting and creative design previously performed by humans.The GenAI-powered solution goes beyond replicating tasks typically handled by third-party agencies and marketing personnelexpanding creative,targeted and personalized marketing in ways that are
198、faster,cheaper and more thorough.“A recent example is the Emmys.Our brands were posting content about the event and related viral moments,”the consumer executive said.“This content was created entirely by GenAI models,picking up some of the trending hashtags,viral clips and news moments,then generat
199、ing a post when it fit with the brand.”He continued,“Of course,we have strong moderation because were putting content out to the public web.We have a human in the loop who monitors content,as well as systems that use reinforcement learning from human feedback.”This highlights thatdespite GenAIs impr
200、essive capabilities and performancehuman engagement is still considered essential to ensure content aligns with brand standards.Case study 3:GenAI is powering an always-on,multimodal social media presence in the consumer industry41Approach The company built on its already strong data and AI foundati
201、on,which included years of experience working with GenAI-related technologies such as natural language processing,cognitive intelligence and multistep reasoning.Over the past 18 months,it has deeply integrated LLMs and foundation models into its business,focusing on architecture,governance and use c
202、ase developmentbalancing build versus buy strategies to maximize impact and value.The companys GenAI strategy has been to rapidly expand and prototype.“In 2023,we were throwing a lot at the wall and seeing what stuck:lots of different providers,architectures,models and experimentation types,”the exe
203、cutive said.“But in 2024,a lot of that coalesced into a strategy weve now codified and defined.”“In this case,our proof of concept took the shape of a pre-GenAI solution we already had that specifically looked at a social media platform analyzing trending influencers and brand affinity.Building on t
204、hat existing dataset,we focused our initial effort on collecting,cleansing,organizing and structuring the data in real time.We then took the data and threw an LLM on top of it to see what kinds of text content it could generate.Later,we expanded our scope to include hashtags,then a multimodal model
205、that includes images,and now short-form video.”Results In the United States,around 60%of the companys brands are using the GenAI-powered solution to achieve an always-on social media presence and produce relevant content with minimal human involvement.The solution is delivering tangible benefits in
206、three key areas.First and foremost is increased productivity,which directly translates into substantial cost savings.“Whether its a first party,second or third party,there were individuals who were conducting these tasks,and there is a dollar value directly associated with each hour of their time,”t
207、he executive said.Second is increased sales,with the GenAI solution helping to boost both the incremental number of impressions for each social media post and the monetary value created by those impressions(due to heightened awareness,increased purchase conviction,and an easier path to purchase).The
208、 third is reduced media costs,particularly the cost savings that accrue when an effective unpromoted social media post eliminates the need to pay for a promoted postfreeing up budget that can be invested more strategically elsewhere.Although many of these benefits have had an immediate impact on the
209、 companys bottom line,some of the productivity gains will take longer to fully realize because they require formal process changes or revisions to existing annual or multiyear contracts.Case study 3:GenAI is powering an always-on,multimodal social media presence in the consumer industry42Authorship
210、and AcknowledgmentsAcknowledgments The authors would like to thank our project sponsors and leaders Nitin Mittal,Kevin Westcott and Jeff Loucks,as well as the additional Deloitte subject matter specialists who contributed to the development of the survey and report:Bjoern Bringmann,Lou DiLorenzo,Roh
211、an Gupta,Kellie Nuttal,Baris Sarer,Ajay Tripathi and Ashish Verma.We would also like to thank our team of professionals who brought this report and campaign to life,including:Ahmed Alibage,Siri Anderson,Hali Austin,Saurabh Bansode,Natasha Buckley,Vanessa Carney,Dystnct Media,Tracy Fulham,Jordan Garr
212、ick,Gerson Lehrman Group(GLG),Karen Hogger,Susie Husted,Lisa Iliff,Wendy Jenkins,Justin Joyner,Diana Kearns-Manolatos,Lena La,Amy Lando,Michael Lim,Cullen Marriott,Rajesh Medisetti,Adriana Mendez,Judy Freeman Mills,Melissa Neumann,Inal Olmez,Jamie Palmeroni,Jonathan Pryce,Negina Rood,Emily Rosenberg
213、,Kate Schmidt,Meredith Schoen,Michael Steinhart,Kelcey Strong,10 EQS,Sandeep Vellanki,Ivana Vucenovic,Talia Wertico,Micah Whitson,Marianne Wilkinson and Sourabh Yaduvanshi.Brenna Sniderman Executive Director Deloitte Center for Integrated Research Deloitte Services LP Jim Rowan Applied AI SGO Leader
214、 Deloitte Consulting LLP Costi Perricos Deloitte Global GenAI Business Leader Deloitte LLP cperricosdeloitte.co.ukBeena Ammanath Executive Director Global Deloitte AI Institute Deloitte LLP Business leadershipResearch leadershipDavid Jarvis Senior Research Leader Deloitte Center for Technology,Media
215、&Telecommunications Deloitte Services LP 43About the Deloitte AI Institute The Deloitte AI Institute helps organizations connect all the different dimensions of the robust,highly dynamic and rapidly evolving AI ecosystem.The AI Institute leads conversations on applied AI innovation across industries
216、,using cutting-edge insights to promote human-machine collaboration in the Age of With.The Deloitte AI Institute aims to promote dialogue about and development of artificial intelligence,stimulate innovation,and examine challenges to AI implementation and ways to address them.The AI Institute collab
217、orates with an ecosystem composed of academic research groups,startups,entrepreneurs,innovators,mature AI product leaders and AI visionaries to explore key areas of artificial intelligence including risks,policies,ethics,future of work and talent,and applied AI use cases.Combined with Deloittes deep
218、 knowledge and experience in artificial intelligence applications,the institute helps make sense of this complex ecosystem and,as a result,delivers impactful perspectives to help organizations succeed by making informed AI decisions.About the Deloitte Center for Integrated ResearchThe Deloitte Cente
219、r for Integrated Research(CIR)offers rigorously researched and data-driven perspectives on critical issues affecting businesses today.We sit at the center of Deloittes industry and functional expertise,combining the leading insights from across our firm to help leaders confidently compete in todays
220、ever-changing marketplace.About the Deloitte Center for Technology,Media&TelecommunicationsThe Deloitte Center for Technology,Media&Telecommunications(TMT Center)is a world-class research organization that serves Deloittes TMT practice and our clients.Our team of professional researchers produce pra
221、ctical foresight,fresh insights,and trustworthy data to help clients see clearly,act decisively and compete with confidence.We create original research using a combination of rigorous methodologies and deep TMT industry knowledge.Learn moreLearn moreLearn more44To obtain a global view of how Generat
222、ive AI is being adopted by organizations on the leading edge of AI,Deloitte surveyed 2,773 leaders between July and September 2024.Respondents were senior leaders in their organizations and included board and C-suite members,and those at the president,vice president and director levels.The survey sa
223、mple was split equally between IT and line of business leaders.Fourteen countries were represented:Australia(100 respondents),Brazil(115 respondents),Canada(175 respondents),France(130 respondents),Germany(150 respondents),India(200 respondents),Italy(75 respondents),Japan(100 respondents),Mexico(10
224、0 respondents),the Netherlands(50 respondents),Singapore(75 respondents),Spain(100 respondents),the United Kingdom(200 respondents),and the United States(1,203 respondents).All participating organizations have one or more working implementations of AI being used daily.Plus,they have pilots in place
225、to explore Generative AI or have one or more working implementations of Generative AI being used daily.Respondents were required to meet one of the following criteria with respect to their organizations AI and data science strategy,investments,implementation approach and value measurement:influence
226、decision-making,are part of a team that makes decisions,are the final decision-maker,or manage or oversee AI technology implementations.All statistics noted in this report and its graphics are derived from Deloittes fourth quarterly survey,conducted July September 2024;The State of Generative AI in
227、the Enterprise:Now decides next,a report series.N(Total leader survey responses)=2,773The survey data was supplemented with case studies and qualitative findings derived from 15 interviews with executives and AI and data science leaders at large organizations across a range of industries.Methodology
228、1.Duncan Stewart,Karthik Ramachandran and Prashant Raman,“Generative AI and cyber:Big risks,but big opportunities too,”Deloitte,November 19,2024,https:/ November 26,2024.2.Emily Mossberg,et al,Global Future of Cyber Survey,4th edition,Deloitte Global,2024,pg 23,https:/ November 26,2024.3.Jeff Loucks
229、,Gillian Crossan,Baris Sarer and China Widener,“Autonomous generative AI agents:Under development,”Deloitte,November 19,2024,https:/ November 26,2024.4.Vivek Kulkarni,Scott Holcomb,Prakul Sharma,Ed Van Buren and Caroline Ritter,“Prompting for action,How AI agents are reshaping the future of work,”De
230、loitte,November 2024,p.16,https:/ November 26,2024.5.“Scaling GenAI:13 Elements for Sustainable Growth and Value,”Deloitte,https:/ November 26,2024.6.Laura Shact,Brad Kreit,Gregory Vert,Jonathan Holdowsky and Natasha Buckley,“Four futures of generative AI in the enterprise:Scenario planning for stra
231、tegic resilience and adaptability,”Deloitte,October 25,2024,https:/ November 26,2024.Endnotes45About DeloitteDeloitte refers to one or more of Deloitte Touche Tohmatsu Limited(DTTL),its global network of member firms,and their related entities(collectively,the“Deloitte organization”).DTTL(also refer
232、red to as“Deloitte Global”)and each of its member firms and related entities are legally separate and independent entities,which cannot obligate or bind each other in respect of third parties.DTTL and each DTTL member firm and related entity is liable only for its own acts and omissions,and not thos
233、e of each other.DTTL does not provide services to clients.Please see to learn more.Deloitte provides industry-leading audit and assurance,tax and related services,consulting,financial advisory,and risk advisory services to nearly 90%of the Fortune Global 500 and thousands of private companies.Our pe
234、ople deliver measurable and lasting results that help reinforce public trust in capital markets,enable clients to transform and thrive,and lead the way toward a stronger economy,a more equitable society,and a sustainable world.Building on its 175-plus year history,Deloitte spans more than 150 countr
235、ies and territories.Learn how Deloittes approximately 457,000 people worldwide make an impact that matters at .This publication contains general information only and Deloitte is not,by means of this publication,rendering accounting,business,financial,investment,legal,tax,or other professional advice
236、 or services.This publication is not a substitute for such professional advice or services,nor should it be used as a basis for any decision or action that may affect your business.Before making any decision or taking any action that may affect your business,you should consult a qualified professional advisor.Deloitte shall not be responsible for any loss sustained by any person who relies on this publication.Copyright 2025 Deloitte Development LLC.All rights