《國際雇主組織&歐盟：2024人工智能（AI）對工作和就業的影響研究報告（英文版）（25頁）.pdf》由會員分享，可在線閱讀，更多相關《國際雇主組織&歐盟：2024人工智能（AI）對工作和就業的影響研究報告（英文版）（25頁）.pdf（25頁珍藏版）》請在三個皮匠報告上搜索。

1、1Analysis of Business Environment in Least Developed CountriesThe Impact of AI on Work and Employment June 2024P O L ICY R E V IE W2This publication was produced with the financial support of the European Union.Its contents are the sole responsibility of the International Organisation of Employers(I

2、OE)and do not necessarily reflect the views of the European Union.3Table of contents I.Factsheet 4II.Introduction 6III.Trends in AI and Employment 8IV.Trends in AI and Labour Management 12V.Policy Challenges and Responses on AI at Work 14VI.Recommendations and Priorities for Employers Organisations

3、and Governments 17VII.References 204Artificial intelligence(AI)refers to systems that can process information,learn from it,and use that learning to generate outputs and achieve goals(Kaplan and Haenlein,2019).Generative artificial intelligence involves using various models to create content such as

4、 text,images,video,or sound.Generative AI has the potential to contribute between$2.6 trillion and$4.4 trillion annually to the global economy,highlighting its positive economic effects(McKinsey&Company,2023).The key impacts of AI on employment are related to job displacement,augmentation,and creati

5、on.While some jobs may be displaced,others will see growth due to AI implementation,particularly in fields like AI modelling and business intelligence(WEF,2023a).There is strong potential for the creation of new jobs,as can be observed over recent years.Generative AI models may increase the value of

6、 jobs requiring social interactions,while the augmentation potential of AI is deemed larger than its automation potential,affecting a wide range of tasks across various job types(Gmyrek,Berg and Bescond,2023).Globally,most CEOs(69 per cent)recognise the need for their workforce to develop new skills

7、 to leverage generative AI effectively(PwC,2024).Among the benefits associated to this technology,AI can enhance productivity across industries,with potential annual global productivity increases of 0.2 per cent to 3.3 per cent(McKinsey&Company,2023).In relation to job quality,while ethical concerns

8、 about AI-based monitoring exist,AI can also improve jobs through reduced tedium,greater engagement,and enhanced safety,exemplified by its use in predicting workplace accidents(Luo et al.,2023)and in personalising training through AI-based virtual tools(Chen,2023).Adoption rates vary,with larger com

9、panies leading the way and smaller ones still in the exploration phase.Globally,it is projected that 74.9 per cent of companies will have adopted AI by 2027,with 59 per cent foreseeing its growing significance in their business strategies(WEF,2023a).AI is increasingly used in hiring processes,perfor

10、mance management,conflict resolution,and downsizing initiatives across various industries.While it offers benefits,ethical concerns regarding data validity,privacy,and bias in decision-making require careful consideration.I.Factsheet5Global AI policies have aimed to establish normative frameworks an

11、d principles.The ILO Centenary Declaration for the Future of Work,OECD AI Principles,UNESCO AI Recommendation are examples.Regional organisations are developing guidelines for AI governance with a focus on human-centred approaches.The EUs AI Act,agreed in December 2023,introduces hard-law regulation

12、 based on risk classification for AI systems,applicable to both public and private entities in the EU market.It will have phased implementation and enforcement by EU countries.Governments have created national AI strategies focused on competitiveness and fairness.Initiatives have concentrated on cre

13、ating institutional capacity,stimulating adoption in public and private sectors,and addressing AI ethics and transparency.Firms have been targeted through direct and indirect support measures such as R&D investment,skills strategies,and data protection legislation(OECD,2024).6II.IntroductionArtifici

14、al intelligence(AI)technologies hold great promise as catalysts for social and economic change in a broader context of digital transformation.AI technologies have the potential to transform businesses,industries,labour markets,and society at large.While there are still obstacles to its application,g

15、enerative AI systems can provide value to customers,optimise processes,supplement human knowledge by providing insights and solutions,and help businesses develop or maintain a competitive advantage.As the need for more strategic involvement and adaptation becomes a reality,employers must have the ne

16、cessary tools and insights to make informed decisions on the right pathways that develop businesses and enhance employment.In this scenario,the purpose of this report is to synthetise existing research about the impact of AI on work and employment,providing strategic knowledge for employers to navig

17、ate its development.While the effects of technology on labour markets are complex and ever-changing,there is a significant body of research that can guide employers decision-making and offer insights into the various dimensions characterising the current process of AI expansion.In this introduction,

18、the report analyses the basic aspects of this technology and the emergence of generative AI in recent years.Section III.responds to key questions about the potential implications of AI for employment with a global focus.Section IV.unpacks some of the current uses of AI in labour management,their pot

19、ential benefits and risks.Section V.analyses the policy landscape of AI with reference to international and national initiatives.Lastly,the report concludes with recommendations for employers and governments regarding AI development and advocacy.1.What is artificial intelligence?Artificial intellige

20、nce(AI)is“a systems ability to interpret external data correctly,to learn from such data,and to use those learnings to achieve specific goals and tasks through flexible adaptation”(Kaplan and Haenlein,2019).The modern conception of AI refers to“agents”or systems that can perform actions based on the

21、ir perception of the environment(Russell and Norvig,2020).Inspired by this conception that sees AI systems as agents,the OECD has defined AI as a machine-based system that,for explicit or implicit objectives,infers how to create outputs from the inputs it receives(e.g.,data),delivering predictions,c

22、ontent,recommendations,or decisions as a result.1 Even though definitions of AI tend to highlight the processing of information and its results,the OECDs definition also emphasises the importance of looking at its objectives,as AI systems can be assessed as beneficial or not based on the aims that a

23、re set for them(Russell,2019).2.What is generative artificial intelligence?Whereas the understanding of AI as“making machines capable of simulating intelligence”is straightforward and uncontroversial(Wamba-Taguimdje et al.,2020),there are recurrent issues in the history of AI that have been subject

24、to societal discussion,such as job displacements,failings in automated systems and privacy protections(Buchanan,2005).Current debates revolve around similar issues,but this time the impact of“generative AI”is in the spotlight.Generative AI uses different models as a foundation to create content in f

25、ormats such as text,images,video or sound.The most resonant of them at present are“large language models”(LLMs)such as Chat-GPT or Gemini.LLMs“can process massive amounts of unstructured text and learn the relationships between words or portions of words”(McKinsey&Company,2023,p.6),allowing them to

26、generate new content that simulate natural language.Generative AI,while displaying great 1 See:https:/oecd.ai/en/wonk/ai-system-definition-update7potential,also faces issues of reliability due to“hallucinations”(incorrect or misleading results that AI models generate)that sound plausible,and copyrig

27、ht concerns.Creators whose data has fed LLM training may challenge the use of their material,potentially exposing users to legal risks(Shanbhogue 2023).Efforts should be made to avoid misinformation as much as possible.Furthermore,many AI researchers say that fakes will become undetectable.An articl

28、e from The Economist talks about fake images,videos and advertisements,etc,as one of the main challenges.The latest software to detect these fakes fail to do so.There is an assumption that AI-generated images or videos will leave a trace.This is not the case.It is getting harder and harder for human

29、s to tell the difference between a real image or video and a fake one.Even the best-performing programme failed to correctly spot computer-generated images 13 per cent of the time(though that was better than the humans,who erred in 39 per cent of cases).Detecting text is slightly better.Watermarking

30、,which is the practice of imperceptibly altering a piece of data in order to embed information about the data,such as tweaking the pixels in subtle ways or shifting their colours,can be useful to enable humans to see the difference.More research is needed on various ways to allow software to detect

31、fake media as this can also impact the world of work.3.What makes generative AI different from previous AI advances?Several factors have made generative AI innovations particularly impactful for the economy.The first of them is that they offer applications in a wide range of activities and industrie

32、s.As a Deloitte report suggests:Apart from its versatility,generative AI has enjoyed widespread use since Chat-GPT was launched in November 2022.The second factor,then,has to do with the high accessibility of current AI-powered applications for content creation,and their reduced cost.A survey from 2

33、022,previous to the release of Chat-GPT,showed that 53 per cent of employers in finance and 58 per cent of them in manufacturing within OECD countries cited the main barrier to AI adoption was the“high costs of technology”(Lane,Williams and Broecke,2023,p.85).Web-based applications that rely on gene

34、rative AI are a game-changer in that sense as they are often free to use or just require a membership.In this regard,generative AI could add between$2.6 trillion to$4.4 trillion to the global economy every year,McKinsey has reported,reinforcing the positive economic potential of this technology(McKi

35、nsey&Company,2023).Around three-fourths of this added value would be about“customer operations,marketing and sales,software engineering,and R&D(McKinsey&Company,2023,p.39).Still,rates of adoption will vary across countries and regions,as well as across enterprises of different sectors and sizes,maki

36、ng its effects spread unevenly(Cazzaniga et al.,2024).“AI took a major leap with Generative AI and its ability to disrupt the way we work because of its ability to create content that profoundly supports human expertise and skillswriting memos and reports,designing website graphics,creating personal

37、ized marketing strategies,and curating employee learning programs,for example.”(Deloitte,2023,p.3)“84.How will AI impact employment in the future?For the past decade,analyses have tended to either highlight labour substitution-the“doomsayers perspective”or the efficiency gains and new types of work

38、derived from AI implementation-the“optimists perspective”(Frank et al.,2019).From a more balanced approach,it is suggested that AI has at least three effects on employment:job displacement,augmentation and creation(Gilbert,2023).Regarding the latter,50 per cent of employers worldwide expect AI to pr

39、omote job growth in the next five years(WEF,2023a,p.6).For employers,the fastest-growing jobs by 2027 will be AI and machine learning specialists,sustainability analysts and business intelligence analysts(WEF,2023a).Apart from increasing demand for already-existing jobs,AI-induced employment growth

40、is expected to be also driven by the emergence of new occupations:prompt engineers,AI modellers,data trainers,as well as governance and ethics specialists are some examples(WEF,2023b).Occupations that require in-person and social interactions will become increasingly valuable,especially with the ris

41、e of generative AI models,as there are“key bottlenecks to the automation of social tasks”(Frey and Osborne,2023,p.3).At the same time,many jobs will be“insulated”from AI innovations,at least in terms of their core skills,for instance in healthcare,construction or hospitality(LinkedIn,2023).Since gen

42、erative AI could have significant implications for a wide range of work-related tasks,it will affect job demand.The augmentation potential is significantly larger than the automation potential:ILO research estimates that while 13 per cent of jobs in the world could be boosted by AI,only 2.3 per cent

43、 could be fully automated at present(Gmyrek,Berg and Bescond,2023,p.34).III.Trends in AI and Employment Prior to the breakthrough of generative AI,analysts had assumed that mainly routine manual and cognitive work was going to be affected by AI and automation(Acemoglu and Autor,2010),predicting that

44、 this was going to create an increasing demand for creative jobs and data analysts.Given the rise of generative AI tools that can excel at both creative and analytic tasks,Deloitte(2023)holds that AI can automate and augment administrative and problem-solving tasks in all types of jobs.For example,i

45、n cognitive jobs(e.g.,accounting or data analysis),in social jobs(e.g.,sales),or physical jobs(e.g.,plumbers or factory work).It is important to recognise,in any case,that projections are based on technical feasibility,available infrastructure,organisational capabilities or incentives to implement A

46、I at work.In general,technology can be a crucial tool for welfare progress,if properly channelled.Simply demonising technology has proved to be the wrong approach.AI can help bring about solutions to fundamental labour and social issues such as those linked to access to healthcare,education,training

47、 or access to social protection.It does not have to be a binary choice of whether to use AI or not;this is a false dichotomy.Furthermore,AI and algorithms are particularly useful for optimising efficiency,including in workforce management.Employers already see such applications being tested in envir

48、onments such as warehouses,distribution,or delivery.AI for managing worker assignments is also useful.However,organisations should ensure this use of AI is done in a responsible way,with respect for workers free choice of completing or improving a task and should,for example,not lead to overly burde

49、nsome workloads for workers.This can subsequently increase the productivity of organisations as well.AI can help minimise or eliminate routine or repetitive tasks.In some instances,policymakers see the use of AI to manage 9assignments as a factor contributing to satisfactory or fulfilling work.Busin

50、esses should engage in conversations with governments about when and where AI can help improve or contribute to solving fundamental labour and social problems.More details on data privacy and data protection can be found below.5.How is AI changing the demand for skills in the labour market?Historica

51、lly,computers have had difficulties performing tasks that humans can easily complete(e.g.,talking or perceiving),while excelling at tasks that involve logic and calculation(Goel and Davies,2011).Nevertheless,in recent years generative AI has rapidly accelerated progress in the technical capabilities

52、 of computers to mimic human characteristics.Whereas in 2017 McKinsey had predicted that skills such as coordination with others,creativity and problem-solving were going to be mastered by computers by 2030 or later.In 2023(post-generative AI developments)McKinsey estimated that by 2030,technology w

53、ill certainly excel at those and several other capacities,such as natural language understanding,but with the exception of social skills(McKinsey&Company,2023).To be sure,this does not necessarily imply that human work will be replaced;rather,it means that employers and employees must have the neces

54、sary skills to use AI tools in their favour and augment or strengthen their own capabilities.The need for upskilling relates,first,to investing in digital skills at different levels:basic(e.g.,accessing computers and smartphones),intermediate(e.g.,job-specific abilities with dedicated software),and

55、advanced(e.g.,programming)(ITU,2021).In relation to advanced digital capacities,the demand for AI skills(i.e.,technical abilities such as software engineering or data analysis)has grown significantly in recent times.Between 2022 and 2023,job applications requiring AI skills increased by 15 per cent

56、in Brazil and 19 per cent in United States(LinkedIn,2023,p.7).On the other hand,demand for social and emotional skills(e.g.,collaboration,conflict resolution,emotional intelligence)is also expected to increase in the near future(Strietska-Ilina and Chun,2021).Certainly,there is awareness in the busi

57、ness community about the upskilling imperative:according to a PwC(2024)global CEO survey,69 per cent of CEOs think that generative AI will ultimately require most of their workforce to develop new skills.New employment challenges and opportunities in algorithmic management can be found below.6.How i

58、s AI making work more productive?Labour productivity can be defined as the average economic output that each individual worker can produce in a certain amount of time.Its levels depend on the technologies workers have at their disposal,their working environment and human capital,among other factors(

59、Cazes and Verick,2012).Evidence suggests that technological innovations play a key role in productivity growth:between 1980 and 2018,40 per cent of variation in labour productivity in emerging markets and developing economies was due to technological change,and in advanced economies it was about 50

60、per cent(Dieppe,2021,p.365).The PwC Global CEO Survey from 2024 shows that 64 per cent of CEOs expect generative AI to increase the efficiency of their employees time at work(PwC,2024).AI is expected to have a similar impact in most industries.As an example,a global study from Massachusetts Institut

61、e of Technology compared workers in customer support using an AI-powered chatbot with others not using it(Brynjolfsson,Li and Raymond,2023).The study concluded that,on average,the former resolved 14 per cent more customer issues than the latter.This benefit was particularly significant for low-skill

62、ed and new workers,who could handle 35 per cent more cases than those not using the chatbot.At an aggregate level,estimates suggest that AI could boost global 10productivity with an annual increase from 0.2 per cent to 3.3 per cent,depending on technology adoption rates(McKinsey&Company,2023,p.44).P

63、roductivity growth,a Goldman Sachs report indicates,will be more noticeable in advanced economies due precisely to the wider adoption of AI and also because of lower productivity growth in these countries.For example,for a 10-year period,it is projected that Japans productivity should grow around 1.

64、5 per cent annually,whereas in India It would be 0.7 per cent and in Colombia 1.1 per cent(Hatzius et al.,2023).7.How is AI affecting the quality of jobs?Given that AI is a general-purpose technology,its potential implications for decent work and job quality can be both positive and negative.Researc

65、h has often echoed a more pessimistic view,suggesting that the use of AI in”algorithmic management”might raise concerns on data protection,privacy and surveillance(OECD 2023),While this form of management can even have counterproductive implications for employers(Giermindl et al.,2022),there is also

66、 evidence that“job quality improvements associated with AI reductions in tedium,greater worker engagement,and improved physical safety may be its strongest endorsement from a worker perspective”(Lane,Williams and Broecke,2023,p.12).For instance,in China,there is use of machine learning methods to pr

67、edict and prevent accidents in the construction sector,based on data about scenarios,equipment and environment(Luo et al.,2023).Companies in OECD countries are also adopting AI tools to automate tedious work,leading to“greater enjoyment on the job”and a greater sense of autonomy(OECD,2023).Likewise,

68、55 per cent of larger companies worldwide that use AI do so for automating repetitive tasks(IBM,2023).Likewise,AI-based virtual trainers,such as applications,can personalise training needs,tailoring the learning process to each individual worker,thus improving their experience and skills acquisition

69、(Chen,2023).In other words,AI can be utilised to improve the working environment and generate better labour conditions.8.How are organisations and companies responding to the development of AI?According to the IBM Global Adoption Index 2023,approximately 42 per cent of enterprise-scale businesses(wi

70、th more than 1,000 workers)say they have actively implemented AI in their operations,while 40 per cent more are still researching or testing AI(IBM,2023).Adoption rates vary by country:IBM also reports that in China,50 per cent of enterprises are actively employing AI,in India 59 per cent and in Lat

71、in America 47 per cent,while larger European economies are still in the exploration phase of AI usage(IBM,2023).Up-to-date data about adoption in small and medium enterprises(SMEs)in developing economies is lacking.Another trend is that large employers are more likely to implement AI:in OECD countri

72、es,around 50 per cent of large manufacturing companies(with more than 500 employees)were using it in 2022,while only 23 per cent of small employers(20 to 49 workers)did(Lane,Williams and Broecke,2023).Adoption of AI is expected to be faster in advanced economies,despite the trend mentioned above in

73、relation to emergent economies,given higher wages and the availability of technology and basic digital infrastructure(McKinsey&Company,2023).Furthermore,projections from the World Economic Forums Future of Jobs Survey(2023)suggests that by 2027,74.9 per cent of companies globally will have adopted A

74、I.Additionally,59 per cent of these companies predict that big data and AI will grow in significance within their business strategies(WEF,2023a,pp.24,46).9.How is the business ecosystem changing due to AI?The transition towards more digitalised business ecosystems is intensifying competition for com

75、panies,with a proliferation of competitors,industries,and commercial models available(Calderon-11Monge and Ribeiro-Soriano,2023).According to a PwC global survey,68 per cent of CEOs believe that“generative AI will increase competitive intensity in my industry”(PwC,2024).In this context,AI is a signi

76、ficant resource,especially for SMEs which necessitate efficient strategies to remain competitive(Kraus et al.,2021).As noted by Drydakis(2022,p.1224),AI applications can be of great value as they”enable SMEs to find new opportunities(sensing capabilities),exploit them(seizing capabilities)and change

77、 operational processes(transforming capabilities)”.Regarding specific applications,it is anticipated that companies will utilise AI across various domains such as process optimisation,customer screening,product and service innovations,and demand forecasting(Lu et al.,2022).AI-enabled tools are offer

78、ing valuable solutions to a series of aspects that provide a competitive advantage to businesses.In relation to logistic barriers and issues in supply chain management,big data analytics can mitigate risk and support knowledge sharing(Zamani et al.,2023).In organisational operations,AI applications

79、allow cross-domain knowledge sharing to maximise the triangulation of data within and with other companies(Enholm et al.,2022).In marketing and sales,AI is used at present for automating e-commerce processes and for customer screening and support via chatbots(Davenport et al.,2020).AI improves servi

80、ce offerings,customer experiences and work efficiency.1210.How is AI being used for hiring employees?Digital recruiting has gone through different phases since the growth of computer use in the 1990s.Initially,it revolved around websites aggregating job offers,which expanded from national to global

81、platforms such as LinkedIn.With the vast number of candidates and offers on these platforms,recruiters have increasingly turned to AI since the 2010s to narrow down opportunities(Black and Van Esch,2020).As of 2023,41 per cent of large enterprises in the world were deploying AI to enhance recruiting

82、 and human resources(HR)processes(IBM,2023).Evidence suggests that HR managers recognise both the advantages and drawbacks of AI usage.On one hand,it enables decision-makers to access a larger pool of candidates,which can be assessed and ranked in a more unbiased manner compared to analogue methods.

83、On the other hand,concerns arise regarding data validity and ethical considerations in candidate data mining(Radonji,Duarte and Pereira,2022;Vrontis et al.,2022).Several tools have been developed to facilitate hiring processes.As Gupta and Mishra(2022)report,there are AI tools that pre-assess candid

84、ates(e.g.,Brazen),screen large numbers of candidate profiles(e.g.,Mya),collect feedback from assessors in different stages of the process(e.g.,Olivia)and provide questionnaires and tests(e.g.,Pymetrics).Additionally,web-scraping applications that infer candidates characteristics based on their socia

85、l media profiles have also been developing over the past decade,raising questions about the ethical implications of AI in recruitment(Dattner et al.,2019).In summary,the literature emphasises the usefulness of AI in processing large amounts of data,while also highlighting IV.Trends in AI and Labour

86、Management the risks that may arise from such practices.The possibilities that AI could bring in terms of labour mobility-increasing the ability to hire personnel based in other countries or who speak different languages-are another promising area to explore through empirical research.11.How is AI b

87、eing used for performance evaluation and reward systems?The utilisation of AI in performance management is spreading fast.According to a global survey of managers,34 per cent of organisations are using AI,such as Chat-GPT,to develop new key performance indicators.Remarkably,90 per cent of these orga

88、nisations have reported improvements in their indicators as a direct result of employing AI(Schrage et al.,2024).In relation to indicator monitoring,AI can be deployed for descriptive,predictive and prescriptive purposes(Leicht-Deobald et al.,2019).Concerning labour management,AI can enhance perform

89、ance monitoring in two ways.Firstly,it can provide more personalised assessments and benchmarks based on individual historical data and contextual factors.Secondly,it can facilitate comparisons across workers through metrics and dashboards(Nyathani,2023).AI-based applications abound:some help manage

90、rs allocate tasks and monitor engagement,while others allow workers access to HR services via chatbots.2 Bias in performance monitoring is also another critical aspect to consider.For instance,Hunkenshroer and Luetge(2022)provide an example of an algorithm that benchmarked top performers in a large

91、company,the majority of whom were males,thus penalising females in future assessment criteria.This profiling or undue discrimination should be carefully avoided.And instead,algorithms should aim to improve services,improve efficiency(time spent working)and productivity.On this point,this should be m

92、ade possible without 2 For a list of examples,see:https:/www.springworks.in/blog/ai-tools-for-hr/13any parties infringing intellectual property rights such as accessing source codes,which are of importance to companies.12.How is AI being used for conflict management in companies?The reliance on AI f

93、or conflict management and resolution is another area of significant development in AI research.It can be utilised for formalising the process of negotiation,such as offering a platform to present preferences and govern the interaction between parties,or in mediation activities where different argum

94、ents have to be unpacked(Aydoan,Baarslag and Gerding,2021).In recent years,AI tools have been developed in the field of dispute resolution(Alessa,2022).Smartsettle ONE is an example,a tool which uses a“blind-bid mechanism”to settle legal disputes and hold negotiation sessions,reducing costs on lawye

95、rs to reach agreements for the parties involved(Financial Times,2019).There is exploratory research on AI-mediated-conversation,which aims to perceive disagreements and propose resolutions,but its application in the workplace has not been extensively studied(Hohenstein and Jung,2020).Even though sys

96、tems of online dispute resolution have improved since the 2000s-for example,through virtual mediation rooms or blind bidding systems(Zeleznikow,2021)-their expansion in businesses worldwide remains an area that requires empirical study.The same can be said about the potential of AI for describing an

97、d predicting conflicts in the workplace.13.How is AI being used for downsizing initiatives?The link between AI and downsizing is twofold:first,it refers to the intentional replacement of human labour by companies,and second,to the utilisation of AI tools to identify redundancies and suggest reductio

98、ns in labour costs.In relation to the first aspect,there is evidence of a reduction in staff costs by automating tasks with AI.In OECD countries,by 2022,around 40 per cent of employers had adopted AI,and this resulted in decreasing labour costs in finance,and in manufacturing this number was closer

99、to 50 per cent(Lane,Williams and Broecke,2023,p.33).Mirroring this trend,more than half of workers in both sectors in OECD economies expressed concern in 2022 about losing their jobs to AI in the next two years(Lane,Williams and Broecke,2023,p.46).There is a dearth of comparative research that exami

100、nes enterprises in developing nations in relation to this matter.In terms of the second aspect-deploying AI tools for layoff decisions-there is evidence of use,but again,there is a lack of comparative data about businesses in developing and emerging economies.In some advanced economies,on the other

101、hand,there is broad use of AI for this purpose.As an example,a Capterra survey of 300 HR leaders in the United States in 2023 showed that in case of an economic recession,98 per cent of them would rely on data and algorithms to make layoff decisions,with 35 per cent saying that they would solely use

102、 data-driven recommendations(Westfall,2023).The study also remarks that the quality and accuracy of companies data,for example on employee skills or performance reviews,is crucial to ensure fairness and efficiency in HR processes.1414.What AI policies have been developed at the international level?T

103、he policy response to AI in global governance during the last decade has primarily involved the elaboration of normative frameworks and principles that governments and companies can adapt to their own AI strategies(Silva 2022).The United Nations has created a High-Level Advisory Body that will deliv

104、er recommendations on the international governance of AI by mid-2024.3 Other relevant international initiatives are:In regional governance,different organisations have proposed soft-law strategies for AI policy,with the common characteristic that all of them have embraced the terminology of a“human-

105、centric”governance approach,also with a focus on the economic value of AI development.For instance,ASEAN(2024)published guidelines for AI governance in member countries in early 2024;in 2023,Latin America and the Caribbean,governments elaborated the Santiago Declaration on ethics of AI(LAC High Leve

106、l Summit,2023);the African Union,on its part,is preparing a Continental AI Strategy(Musoni,2024).V.Policy Challenges and Responses on AI at Work ILO Centenary Declaration for the Future of Work(2019):promotes developing a human-centred approach that focuses on strengthening peoples capacities,reinfo

107、rcing the institutions of work and promoting sustained,inclusive and sustainable economic growth and decent work.OECD AI Principles(2019):supports human-centred values like respecting the rule of law,human rights,and democratic values throughout the AI lifecycle,aiming to promote inclusive growth an

108、d transparent AI development.UNESCO Recommendation on AI(2021):emphasises the significance of human supervision of AI systems,and that policymakers should work alongside representatives from the private sector,civil society,unions and other stakeholders for a fair transition.G7 Hiroshima AI Internat

109、ional Guiding Principles and Code of Conduct(2023):promotes the safe and responsible development of AI systems through risk mitigation,transparency,governance,security,research prioritisation,and adoption of standards.3 See:https:/www.un.org/techenvoy/ai-advisory-body.The UN General Assembly,indicat

110、ing the direction where the Advisory Body might go in terms of AI governance,adopted a resolution in March 2023 that“called on all Member States and stakeholders to refrain from or cease the use of artificial intelligence systems that are impossible to operate in compliance with international human

111、rights law or that pose undue risks to the enjoyment of human rights”(UN 2024).15Some areas need to be thoroughly analysed and considered to release all the potentiality of AI in the World of work:a.Job losses and job creation.It is likely that AI,as can be expected with any disruptive technological

112、 change,will bring about job losses and job creation.The key policy challenge is to ensure successful transitional stories of job and income opportunities for more individuals globally.Effective transitions mean that there is a need to have open and dynamic labour markets.It also means that there is

113、 a need for efficient and effective educational and lifelong learning systems to respond to current and future skill needs.Enhancing a“learnability”attitude(i.e.the capacity to learn quickly),will be more crucial than ever,as well as efficient labour market institutions facilitating the shift of ski

114、lls development needs.Skills are both challenged and augmented by AI,and“Meta-Learning”(Learning how to learn)is more critical than ever(Fadel,2024).b.Algorithmic management.Policymakers are concerned with potential abusive behaviour when designing,developing or launching algorithms for the purposes

115、 of work-related monitoring systems.These systems are meant to promote productivity increases and even enhance work-life balance.Some analysts warn about the need to prevent inappropriate discrimination,the need to comply with working time regulations,and the need to balance productivity with monito

116、ring and decent working conditions.Some solutions are already in place to settle an appropriate algorithm management,and this will likely be an area for policy action.On this point,maintaining confidentiality in the algorithm formulas is essential to preserve intellectual property rights,with some d

117、egree of transparency to be implemented.c.Cybersecurity at the workplace and workers data.With AI expanding,cybersecurity and data management at the workplace have become a more challenging and relevant area of potential policy action.There will likely be a need to avoid or mitigate massive AI attac

118、ks on sensitive data affecting workers privacy or Human Resources confidential information.This point is also linked to how AI uses workers private data to guide advice on Human Resources policies.d.Misinformation at the workplace.As mentioned above,the wrong use of AI can also lead to misinformatio

119、n affecting the workplace(fake certificates,fake voices,fake information on workers behaviours or private lives).Again,this represents a growing area of policy action.4 The EU AI Act distinguishes between“minimal risk”systems(e.g.,recommender systems or spam filters),“high risk”(e.g.,critical infras

120、tructures;medical devices;recruitment systems),“unacceptable risk”(e.g.,social scoring,predictive policing,emotion recognition),and“specific transparency risk”(e.g.,chatbots or deepfakes).Among these,minimal risk systems will be allowed to operate.High-risk ones will need to specify risk-mitigation

121、systems to regulators.Unacceptable risk systems will not be allowed to operate.Lastly,specific transparency systems will require that contents have an explicit recognition that they have been generated by an AI.Non-compliance will be fined with a 7 per cent of total global annual turnover,and supply

122、 of incorrect or misleading information will entail a 1.5 per cent fine.Guidance mentions that in principle SMEs would have lower fines than larger companies.For more information,see European Commission(2023a).1615.What does the European Unions AI Act entail for business?In December 2023,the Europea

123、n Union(EU)reached an agreement with members on the first hard-law regulation:the EU AI Act.The Act takes a risk-based approach to regulation by classifying AI systems according to their risk profile.4 In terms of its implications for businesses:Accordingly,EU countries,in collaboration with new Eur

124、opean AI Office,will be in charge of assessing AI applications.Some of the AI tools that are being utilised by companies at present could be subject to scrutiny as a result:for instance,emotion recognition in the workplace is deemed“unacceptable”by the Act,AI-powered recruitment techniques could be

125、classified as“high risk”,and generative AIs such as Chat-GPT,might not face major constraints insofar as they can ensure transparency for users(European Commission,2023b).The timeline of the AI Act implies that 6 months after entering into force,EU states must start phasing out“prohibited systems”an

126、d after 24 months,“all rules would become applicable”,including regulation of high-risk systems(European Commission,2023a).16.What are the policy responses and priorities at the domestic level?The OECD AI Policy Observatory serves as a vital data source for analysing national AI policies across most

127、 countries globally.5 Most initiatives have consisted in the creation of national strategies for AI development,in evaluating regulations and governance instruments.Firms have been an important target group for government initiatives.Among the 407 initiatives the OECD reports about,115 have been foc

128、used on SMEs,43 on multinational companies,40 on large firms and 25 on micro-enterprises.6 Within the large variety of initiatives that can be found globally,national AI governance can be broadly divided into three areas:1.Creation of institutional capacity:new institutions have been created to unif

129、y government efforts and initiatives related to AI and to provide advice for policymakers.The institutional landscape of AI varies from country to country(Radu,2021).2.Stimulating AI adoption:governments are promoting adoption in both the public and private sectors.Direct government support includes

130、 providing grants for AI project development in companies and funding for AI startups.Indirect support encompasses policies such as skills strategies aimed at promoting AI skills within education systems,as well as the utilisation of tax credits related to R&D investment in companies(Milanez,2023).A

131、dditionally,government-funded R&D investment has promoted national AI research centres and networks(Galindo,Perset and Sheeka,2021).3.AI ethics and transparency:governments are aiming to balance the potential benefits of AI with consideration of issues that might undermine the trustworthiness of AI.

132、The development of policies in that line has concentrated on creating new ethical advice bodies and elaborating normative frameworks that can be used as guidelines for AI adoption in each country,which tend to be voluntary in terms of enforcement(OECD,2024).Another focal point of initiatives has bee

133、n data protection legislation;most countries already have regulations in place,but some have created new norms to address digital privacy concerns(UNCTAD,2021).5 See:https:/oecd.ai/en/dashboards/overview6 As of February 2024.See:https:/oecd.ai/en/dashboards/overview/target17Recommendations for Emplo

134、yers Organisations and CompaniesThe priorities of companies and employer organisations regarding the emergence of AI can be categorised into two levels.Firstly,individual organisations need to take proactive steps to leverage the benefits of this technology while mitigating various risks.Secondly,bu

135、siness advocates must concentrate their efforts on promoting AI adoption that enhances productivity in collaboration with governments.Organisation-level recommendations:Promote a culture of innovation:where AI adoption is deemed a feasible option and appropriate governance systems are in place,organ

136、isations should encourage a culture of innovation that allows for practical experimentation and learning from successes or failures.This will support strategy execution but also ensure that organisations stay informed about AI developments and are better prepared to pivot or adapt strategies as need

137、ed.Elaborate an AI strategy for your organisation:AI can improve business performance and innovativeness,but employers need first to elaborate company-specific AI strategies that serve as a navigation chart in their journey.It can be a similar approach to how companies have dealt with technologies i

138、n the past(such as when the internet was first developed).At the global level,in 2023 59 per cent of large enterprises had an AI strategy in place,while 32 per cent was developing one,and only 4 per cent did not have any(IBM,2023,p.22).Although data on SMEs is still scarce,research suggest that they

139、 need to plan ahead on three levels of AI capability:the creation of organisational resources,data resources and responsible AI governance mechanisms to ensure the efficiency and fairness in their structural changes(Enholm et al.,2022).Investing in knowledge and skills development:These must be a pr

140、iority for business in promoting job productivity.Some need to develop an understanding(knowledge),and others will need to be able to apply this in a practical way(skills).An Adecco global survey shows that only 46 per cent of workers were receiving guidance on how to use AI in their tasks in 2023,i

141、ndicating that this remains a significant challenge for business AI strategies(Adecco Group 2023).AI can act as a co-pilot where it can assist humans to be in control.AI is created by people,for people.Evaluate ethical risk assessments of AI tools:given the importance of ethical risk in public and p

142、olicy discourse,companies should approach this dimension with seriousness.This involves assessing issues related to data availability and access,the internal governance of AI in the workplace,and normative aspects that influence employers interactions with workers and customers(Lu et al.,2022).In ot

143、her words,“further deliberation is needed to address governance issues such as the handling of personal information included in AI training data and the disclosure of personal information in generative AI outputs”(Keidanren 2023).Social dialogue and collaboration with governments and workers is a vi

144、able strategy to face ethical concerns regarding AI adoption,and it can facilitate the introduction of AI innovations and their design at work(OECD 2023).Implement transition programmes:this can be in the form of retraining and redeployment initiatives.VI.Recommendations and Priorities for Employers

145、 Organisations and Governments18Advocacy-level recommendations:Promote research or the collection of data for SMEs or in developing countries to support decision-making,strategy or policy development:as mentioned above,there were references to gaps in available data for SMEs in developing countries.

146、Employers can also advocate for sector-specific guidelines,recognising that AIs value,impact and risks may vary across sectors.Promote a smooth transition for AI adoption:as mentioned above,AI has at least three effects on employment:job displacement,augmentation and creation.What matters most is pr

147、omoting a smooth transition to minimise net job losses.It can also help move businesses and workers from routine to more creative jobs.Transitions mean open and dynamic labour markets.Robust educational and training systems are needed to respond to real skills needs.This is also linked to having an

148、open mindset to learning and the capacity to learn quickly.Strong institutions can help facilitate the shift of skills development.Otherwise,countries and companies will face massive disruptions in the labour market,perhaps more than ever.Promote policies that address barriers to AI adoption:current

149、ly,the most prominent obstacles to AI development faced by employers are linked to a shortage of AI skills and expertise,an inability to manage complex datasets,and integrate and scale them within organisational operations(IBM,2023).Employers can advocate for government policies through direct and i

150、ndirect support initiatives.In particular,reskilling and upskilling programmes are the type of measures that most employers see as beneficial to increase their capacity for technological innovation(WEF,2023a).At the same time,supporting social safety nets might help to make this transition more soci

151、ally sustainable(Cazzaniga et al.,2024).Policymakers should not stifle innovation as AI and other forms of technology can assist micro,small and medium sized companies,offer business and economic opportunities for entrepreneurs and those in the world of work.New regulations are not necessarily the p

152、anacea to resolve issues,it is a matter to embedding a culture of ethical awareness and a mindset of adaptation.Establish partnerships with stakeholders to support AI adoption:the diffusion of innovations in national economies will probably come from a synergy of actors supporting AI implementation.

153、Partnerships with international organisations,development agencies,experts,civil society and workers organisations might catalyse this process.For instance,university collaborations can help establish connections between governments,academic research and industry,facilitating knowledge sharing and t

154、raining opportunities for companies and also increasing AI literacy in the business ecosystem(Milanez,2023).This could mean a new set of partnerships unseen before.Policy Recommendations for Governments General risk mitigation should be given more prominence:particularly in respect of conducting per

155、iodic risk assessments and developing appropriate contingency plans,including investments in suitable cybersecurity infrastructure.At the government level,promoting investments in cybersecurity infrastructure,particularly for SMEs,may be a useful consideration.An example can be observed in Trinidad

156、and Tobago:the government introduced a Cybersecurity Investment Tax Allowance in January 2024,which offers a tax deduction of up to$500,000 TTD(approx.74,000 USD)to businesses that invest in eligible cybersecurity software and network security monitoring equipment.Issues such as data protection and

157、security of workers data and sensitive 19company data(HR and financial),and use of IT devices,equipment,software and services should be reviewed and monitored.There are various cases where it is easy to impersonate a person at the workplace.Governments also have a responsibility,in terms of facilita

158、ting digital literacy programmes across all levels of society.This will ensure that its citizens are better prepared for the realities of working in an AI-based workplace or economy.Providing unemployment relief support(social protection)or other types of retraining or employment programmes for work

159、ers who are in transition.20VII.ReferencesAcemoglu,D.and Autor,D.(2010)Skills,Tasks and Technologies:Implications for Employment and Earnings.w16082.Cambridge,MA:National Bureau of Economic Research,p.w16082.Available at:https:/doi.org/10.3386/w16082.Adecco Group(2023)Whats working?Navigating the AI

160、 revolution and the shifting future of work.Available at:https:/ role of Artificial Intelligence in Online Dispute Resolution:A brief and critical overview,Information&Communications Technology Law,31(3),pp.319342.Available at:https:/doi.org/10.1080/13600834.2022.2088060.ASEAN(2024)ASEAN Guide on AI

161、 Governance and Ethics.Available at:https:/asean.org/book/asean-guide-on-ai-governance-and-ethics/Aydoan,R.,Baarslag,T.and Gerding,E.(2021)Artificial Intelligence Techniques for Conflict Resolution,Group Decision and Negotiation,30(4),pp.879883.Available at:https:/doi.org/10.1007/s10726-021-09738-x.

162、Black,J.S.and Van Esch,P.(2020)AI-enabled recruiting:What is it and how should a manager use it?,Business Horizons,63(2),pp.215226.Available at:https:/doi.org/10.1016/j.bushor.2019.12.001.Brynjolfsson,E.,Li,D.and Raymond,L.(2023)Generative AI at Work.w31161.Cambridge,MA:National Bureau of Economic R

163、esearch,p.w31161.Available at:https:/doi.org/10.3386/w31161.Buchanan,B.(2005)A(Very)Brief History of Artificial Intelligence,AI Magazine,26(4).Available at:https:/doi.org/10.1609/aimag.v26i4.1848.Calderon-Monge,E.and Ribeiro-Soriano,D.(2023)The role of digitalization in business and management:a sys

164、tematic literature review,Review of Managerial Science Preprint.Available at:https:/doi.org/10.1007/s11846-023-00647-8.Cazes,S.and Verick,S.(eds)(2012)Perspectives on labour economics for development.Geneva:International Labour Office.Cazzaniga,M.et al.(2024)Gen-AI:Artificial Intelligence and the Fu

165、ture of Work,IMF Staff Discussion Notes,2024(001),p.1.Available at:https:/doi.org/10.5089/9798400262548.006.Chen,Z.(2023)Artificial Intelligence-Virtual Trainer:Innovative Didactics Aimed at Personalized Training Needs,Journal of the Knowledge Economy,14(2),pp.20072025.Available at:https:/doi.org/10

166、.1007/s13132-022-00985-0.Dattner,B.et al.(2019)The Legal and Ethical Implications of Using AI in Hiring,Harvard Business Review,25 April.Available at:https:/hbr.org/2019/04/the-legal-and-ethical-implications-of-using-ai-in-hiring.Davenport,T.et al.(2020)How artificial intelligence will change the fu

167、ture of marketing,21Journal of the Academy of Marketing Science,48(1),pp.2442.Available at:https:/doi.org/10.1007/s11747-019-00696-0.Deloitte(2023)Generative AI and the future of work.Deloitte.Available at:https:/ productivity:trends,drivers,and policies.Washington,DC:World Bank Group.Economist(2024

168、)Many AI researchers think fakes will become undetectable.Available at:https:/ al.(2022)Artificial Intelligence and Business Value:a Literature Review,Information Systems Frontiers,24(5),pp.17091734.Available at:https:/doi.org/10.1007/s10796-021-10186-w.European Commission(2023a)Artificial intellige

169、nce-questions and answers.Available at:https:/ec.europa.eu/commission/presscorner/detail/en/QANDA_21_1683.European Commission(2023b)Commission welcomes political agreement on Artificial Intelligence Act.Available at:https:/ec.europa.eu/commission/presscorner/detail/en/ip_23_6473.Fadel,C.et al.(2024)

170、Education for the Age of AI.Available at:https:/curriculumredesign.org/our-work/education-for-the-age-of-ai/.Financial Times(2019)Robots and AI threaten to mediate disputes better than lawyers,14 August.Available at:https:/ al.(2019)Toward understanding the impact of artificial intelligence on labor

171、,Proceedings of the National Academy of Sciences,116(14),pp.65316539.Available at:https:/doi.org/10.1073/pnas.1900949116.Frey,C.and Osborne,M.(2023)Generative AI and the Future of Work:A Reappraisal.Working Paper No.2023.Oxford Martin School.G7(2023)Hiroshima Process International Code of Conduct fo

172、r Organizations Developing Advanced AI Systems.G7 2023 Hiroshima Summit.Galindo,L.,Perset,K.and Sheeka,F.(2021)An overview of national AI strategies and policies.Going Digital Toolkit Note 14.Available at:https:/goingdigital.oecd.org/data/notes/No14_ToolkitNote_AIStrategies.pdfGiermindl,L.M.et al.(2

173、022)The dark sides of people analytics:reviewing the perils for organisations and employees,European Journal of Information Systems,31(3),pp.410435.Available at:https:/doi.org/10.1080/0960085X.2021.1927213.Gilbert,A.(2023)Reframing Automation-a new model for anticipating risks and impacts.Institute

174、for the Future of Work(UK).Available at:https:/zenodo.org/record/8099822(Accessed:18 February 2024).22Gmyrek,P.,Berg,J.and Bescond,D.(2023)Generative AI and Jobs:A Global Analysis of Potential Effects on Job Quantity and Quality,SSRN Electronic Journal,ILO Working Paper 96.Available at:https:/doi.or

175、g/10.2139/ssrn.4584219.Goel,A.K.and Davies,J.(2011)Artificial Intelligence,in R.J.Sternberg and S.B.Kaufman(eds)The Cambridge Handbook of Intelligence.1st edn.Cambridge University Press,pp.468482.Available at:https:/doi.org/10.1017/CBO9780511977244.024.Gupta,A.and Mishra,M.(2022)Ethical Concerns Whi

176、le Using Artificial Intelligence in Recruitment of Employees,Business Ethics and Leadership,6(2),pp.611.Available at:https:/doi.org/10.21272/bel.6(2).6-11.2022.Hatzius,J.et al.(2023)The Potentially Large Effects of Artificial Intelligence on Economic Growth.Goldman Sachs.Available at:https:/ Jung,M.

177、(2020)AI as a moral crumple zone:The effects of AI-mediated communication on attribution and trust,Computers in Human Behavior,106,p.106190.Available at:https:/doi.org/10.1016/j.chb.2019.106190.Hunkenschroer,A.L.and Luetge,C.(2022)Ethics of AI-Enabled Recruiting and Selection:A Review and Research A

178、genda,Journal of Business Ethics,178(4),pp.9771007.Available at:https:/doi.org/10.1007/s10551-022-05049-6.IBM(2023)Global AI Adoption Index 2023.IBM.Available at:https:/ Declaration for the Future of Work.Geneva:International Labour Office.ITU(2021)Digital skills insights 2021.Geneva:International T

179、elecommunication Union.Kaplan,A.and Haenlein,M.(2019)Siri,Siri,in my hand:Whos the fairest in the land?On the interpretations,illustrations,and implications of artificial intelligence,Business Horizons,62(1),pp.1525.Available at:https:/doi.org/10.1016/j.bushor.2018.08.004.Kraus,S.et al.(2021)Digital

180、 Transformation:An Overview of the Current State of the Art of Research,SAGE Open,11(3),p.215824402110475.Available at:https:/doi.org/10.1177/21582440211047576.LAC High Level Summit(2023)Declaracion de Santiago.Para promover una inteligencia artificial tica en Amrica Latina y el Caribe.Available at:

181、https:/minciencia.gob.cl/uploads/filer_public/40/2a/402a35a0-1222-4dab-b090-5c81bbf34237/declaracion_de_santiago.pdfLane,M.,Williams,M.and Broecke,S.(2023)The impact of AI on the workplace:Main findings from the OECD AI surveys of employers and workers.OECD Social,Employment and Migration Working Pa

182、pers 288.Available at:https:/doi.org/10.1787/ea0a0fe1-en.Leicht-Deobald,U.et al.(2019)The Challenges of Algorithm-Based HR Decision-Making for Personal Integrity,Journal of Business Ethics,160(2),pp.377392.Available at:https:/doi.org/10.1007/s10551-019-04204-w.23Linkedin(2023)Future of work report.A

183、I at work.Available at:https:/ al.(2022)AI-Enabled Opportunities and Transformation Challenges for SMEs in the Post-pandemic Era:A Review and Research Agenda,Frontiers in Public Health,10,p.885067.Available at:https:/doi.org/10.3389/fpubh.2022.885067.Luo,X.et al.(2023)Application of machine learning

184、 technology for occupational accident severity prediction in the case of construction collapse accidents,Safety Science,163,p.106138.Available at:https:/doi.org/10.1016/j.ssci.2023.106138.McKinsey&Company(2023)The economic potential of generative AI:The next productivity frontier.McKinsey.Available

185、at:https:/ impact of AI on the workplace:Evidence from OECD case studies of AI implementation.OECD Social,Employment and Migration Working Papers 289.Available at:https:/www.oecd.org/publications/the-impact-of-ai-on-the-workplace-evidence-from-oecd-case-studies-of-ai-implementation-2247ce58-en.htm(A

186、ccessed:19 February 2024).Musoni,M.(2024)Envisioning Africas AI governance landscape in 2024,ECDPM,January.Available at:https:/ecdpm.org/work/envisioning-africas-ai-governance-landscape-2024.Nyathani,R.(2023)AI in Performance Management:Redefining Performance Appraisals in the Digital Age,Journal of

187、 Artificial Intelligence&Cloud Computing,pp.15.Available at:https:/doi.org/10.47363/JAICC/2023(2)134.OECD(2019)OECD AI Principles.Available at:https:/oecd.ai/en/ai-principles/.OECD(2023)OECD Employment Outlook 2023:Artificial Intelligence and the Labour Market.Available at:https:/doi.org/10.1787/087

188、85bba-en.OECD(2024)OECD AI Policy Observatory:Regulatory oversight and ethical advice bodies.Available at:https:/oecd.ai/en/dashboards/policy-instruments/Regulatory_oversight_and_ethical_advice_bodies.PwC(2024)Thriving in an age of continuous reinvention.PwC.Available at:https:/ Pereira,N.(2022)Arti

189、ficial intelligence and HRM:HR managers perspective on decisiveness and challenges,European Management Journal,p.S0263237322000883.Available at:https:/doi.org/10.1016/j.emj.2022.07.001.Radu,R.(2021)Steering the governance of artificial intelligence:national strategies in perspective,Policy and Socie

190、ty,40(2),pp.178193.Available at:https:/doi.org/10.1080/14494035.2021.1929728.Russell,S.and Norvig,P.(2020)Artificial intelligence:A modern approach.4.ed.London:Pearson.Russell,S.J.(2019)Human compatible:artificial intelligence and the problem of control.New York.Viking.24Shanbhogue,R.(2023).Generati

191、ve AI holds much promise for businesses.The Economist.Available at:https:/ al.(2024)The Future of Strategic Measurement:Enhancing KPIs with AI.MIT Sloan Management Review.Available at:https:/shop.sloanreview.mit.edu/store/the-future-of-strategic-measurement-enhancing-kpis-with-ai.Silva,V.(2022).The

192、Schumpeterian consensus:The new logic of global social policy to face digital transformation.Journal of Social Policy,pp.117.Available at:https:/doi.org/10.1017/s0047279422000861.Strietska-Ilina,O.and Chun,H.-K.(2021)Changing demand for skills in digital economies and societies:literature review and

193、 case studies from low-and middle-income countries.Available at:https:/www.ilo.org/publications/changing-demand-skills-digital-economies-and-societies-literature-review.UN(2024).General Assembly adopts landmark resolution on artificial intelligence.UN News.Available at:https:/news.un.org/en/story/20

194、24/03/1147831.UNCTAD(2021)Data Protection and Privacy Legislation Worldwide.Available at:https:/unctad.org/page/data-protection-and-privacy-legislation-worldwide.UNESCO(2021)UNESCO Recommendation on the Ethics of AI.Available at:https:/www.unesco.org/en/articles/recommendation-ethics-artificial-inte

195、lligence.Vrontis,D.et al.(2022)Artificial intelligence,robotics,advanced technologies and human resource management:a systematic review,The International Journal of Human Resource Management,33(6),pp.12371266.Available at:https:/doi.org/10.1080/09585192.2020.1871398.Wamba-Taguimdje,S.-L.et al.(2020)

196、Influence of artificial intelligence(AI)on firm performance:the business value of AI-based transformation projects,Business Process Management Journal,26(7),pp.18931924.Available at:https:/doi.org/10.1108/BPMJ-10-2019-0411.WEF(2023a)Future of jobs report 2023.World Economic Forum.Available at:https:

197、/www3.weforum.org/docs/WEF_Future_of_Jobs_2023.pdf.WEF(2023b)Jobs of Tomorrow:Large Language Models and Jobs.World Economic Forum.Available at:https:/www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf.Westfall,B.(2023)Algorithms Will Make Critical Talent Decisions in the Next Recessio

198、n.Capterra.Available at:https:/ al.(2023)Artificial intelligence and big data analytics for supply chain resilience:a systematic literature review,Annals of Operations Research,327(2),pp.605632.Available at:https:/doi.org/10.1007/s10479-022-04983-y.Zeleznikow,J.(2021)Using Artificial Intelligence to provide Intelligent Dispute Resolution Support,Group Decision and Negotiation,30(4),pp.789812.Available at:https:/doi.org/10.1007/s10726-021-09734-1.25 IOE 2024Avenue Louis-Casa 71 CH-1216 GenveT+41 22 929 00 00 F+41 22 929 00 01 ioeioe- ioe-emp.org