Guidelines for AI and Shared Prosperity

Step 2: Apply the Job Impact Assessment Tool

Step 2: Apply the Job Impact Assessment Tool

Use the high-level Job Impact Assessment Tool to analyze a given AI system:

Go over the full list of signals of opportunity and risk

Analyze the distribution of potential benefits and harms

Repeat this process for upstream and downstream markets

Instructions for Performing a Job Impact Assessment

Instructions for Performing a Job Impact Assessment

Assess the AI system against the full list of signals
Go over the full list of signals of opportunity and risk and document which signals are present in the case of the AI system being assessed. Not all signals apply for every AI system. Document those that do not apply as not applicable, but do not skip or cherry-pick signals. For each step, document the explanation for the answer for future reference.

For each signal, if you estimated the likelihood of the respective opportunity or risk materializing as a result of the introduction of the AI system into the economy to be anything but “zero,” please note the respective signal as “present.”

Certainty in likelihood estimation is not a prerequisite for this high-level assessment and is assumed to be absent in most cases. When in doubt, note the signal as “present.”

Analyze the distribution of potential benefits and harms
Document in as much detail as possible your understanding of the distribution of potential benefits and harms of an AI system across skill, geographic, and demographic groups, and how it might change over time.✱Relevant time period depends on how long the AI system being assessed is expected to remain in use. (Are today’s “winners” expected to lose their gains in the future? The reverse?) The exact steps needed to perform the distribution of impacts analysis are highly case-specific. PAI is looking to engage with stakeholders to curate a library of distribution analysis examples for the community to learn from. If you would like to contribute to this, please get in touch.

GET IN TOUCH

Repeat this process for upstream and downstream markets
In order to take into account the possible effects on the competitors, suppliers, and clients of the AI-using organization, repeat the signal detection and analysis processes not only for the primary market the AI system is intended to be deployed in, but also upstream and downstream markets.
Proceed to our Stakeholder-Specific Recommendations
After completing the high-level Job Impact Assessment analysis, AI-creating and AI-using organizations should implement recommended Responsible Practices (where not already in use) to improve anticipated outcomes — for instance, to eliminate or mitigate anticipated harms or increase likely benefits for workers and the economy. These Responsible Practices can be found under Step 3 of the Shared Prosperity Guidelines. (Responsible Practices will be added and refined through community testing and feedback.)

Policymakers, workers and their representatives can use the results of the high-level Jobs Impact Assessment to inform their decisions, actions, and agendas as outlined in the Suggested Uses section under Step 3 of the Shared Prosperity Guidelines. We look forward to collecting feedback on the Guidelines and curating use examples in partnership with interested stakeholders. To get involved, please get in touch.

GET IN TOUCH

Signals of Opportunity to Advance Shared Prosperity

Signals of Opportunity to Advance Shared Prosperity

If one or more of the statements below apply to the AI system being assessed, this indicates a possibility of a positive impact on shared prosperity-relevant outcomes.

An opportunity signal (OS) is present if an AI system may:

OS1: Generate significant, widely distributed benefits
Will the AI system generate significant, widely distributed benefits to the planet, the public, or individual consumers? One of the primary motivations for investing in the research and development of AI is its potential to help humanity overcome some of our most pressing challenges, including ones related to climate change and the treatment of disease. Hence, the potential of an AI system to generate public goods or benefit the environment is a strong signal of opportunity to advance shared prosperity.Individual consumer benefits can be more controversial as many advocates point out the growing environmental costs that frequently accompany the commodification of consumer goods. But if production and consumption are environmentally conscious, a potential to generate significant and widely distributed consumer benefits is a signal of opportunity to advance shared prosperity. Cheaper or more high-quality goods or services make consumers richer in real terms,✱This is a result of the “real income effect.” For the same nominal amount of money, consumers are able to buy more or higher quality goods. freeing up parts of their incomes to be spent to buy other goods and services, boosting the demand for labor in respective sectors of the economy.

How significant and widely distributed consumer benefits should be to justify job losses is a political question,✱For example, in 2011, the US government imposed tariffs to prevent job losses in the tire industry. Economic analysis later showed that the tariffs cost American consumers around $0.9 million per job saved: https://www.piie.com/publications/policy-briefs/us-tire-tariffs-saving-few-jobs-high-cost. It seems implausible that such large consumer costs are worthwhile, relative to the job gains. but quantifying consumer gains per job lost would help sharpen up any debate about the value of an AI innovation.✱In this paper, Brynjolfsson et al. estimate the value of many free digital goods and services: https://www.pnas.org/doi/10.1073/pnas.1815663116. They do so by proposing a new metric called GDP-B, which quantifies their benefits rather than costs, and then estimating consumers’ willingness-to-pay for free digital goods and services in terms of GDP-B.Brynjolfsson, E., Collis, A., Diewert, W.E., Eggers, F., and Fox, K.J. (2019). GDP-B: Accounting for the value of new and free goods in the digital economy (No. w25695). National Bureau of Economic Research. As stated in “Key Principles for Using the Guidelines,” independently of the magnitude and distribution of anticipated benefits, appropriate mitigation strategies should be developed in response to the risk of job losses or wage decreases.

OS2. Boost worker productivity

Will the AI system boost productivity of workers, in particular those in lower-paid jobs, without increasing strain? By a worker’s productivity, we mean a worker’s output per hour. A more productive worker is more valuable to their employer and (all other conditions remaining the same) is expected to be paid more.✱As emphasized in Key Principles for Using the Guidelines, signals of opportunity are not guarantees: It is possible that the introduction of a new technology into the workplace boosts workers’ productivity but does not lead to wage growth because, in practice, workers’ productivity is only one of the factors determining their wage. Other factors include how competitive the market is and how much bargaining power workers have. In fact, a large number of countries have been experiencing productivity-wage decoupling in recent decades (see, for example: https://www.oecd.org/economy/decoupling-of-wages-from-productivity/). This points to a diminishing role of productivity in determining wages, but it remains non-zero and hence has to be accounted for by the Guidelines. Therefore, if an AI system comes with a promise of a productivity boost that is a positive signal. Besides, productivity growth is often the prerequisite for the creation of consumer benefits discussed in OS1.However, three important caveats should be noted here.

Caveat 1: Productivity boosts can deepen inequality

It is quite rare for a technology to equally boost productivity for everyone involved in the production of a certain good, more often it helps workers in certain skill groups more than others. If it is helping workers in lower-paying jobs relatively more, the effect could be inequality-reducing. Otherwise, it may be inequality-deepening. Please document the distribution of the productivity increase across the labor force when assessing the presence of this opportunity signal.

Caveat 2: Productivity boosts can displace workers

Even if productivity of all workers involved in the production of a certain good is boosted equally by an AI system, fewer of them might find themselves employed in the production of that good once the AI system is in place. This is because fewer (newly more productive) worker-hours✱The impact of a productivity-enhancing technology can manifest itself as a reduction of the size of the workforce, or a reduction in hours worked by the same-size labor force. Either option can negatively impact shared prosperity. are now needed to create the same volume of output. For production of the good in question to require more human labor after AI deployment, two conditions must be met:

  • Productivity gains of the firm introducing AI need to be shared with its clients (such as consumers, businesses, or governments) in the form of lower-priced or higher-quality products — something which is less likely to happen in a monopolistic environment
  • Clients should be willing to buy sufficiently more of that lower-priced or higher-quality product

If the first condition is met but the second is not, the introduction of the AI system in question might still be, on balance, labor-demand boosting if it induces a “productivity effect” in the broader economy. When productivity gains and corresponding consumer benefits are sufficiently large, consumers will experience a real income boost generating new labor demand in the production of complementary goods. That new labor demand might be sufficient to compensate for the original loss of employment due to an introduction of an AI system. Issues arise when the productivity gains are too small like in the case of “so-so” technologiesAcemoglu, D., and Restrepo, P. (2019). Automation and new tasks: How technology displaces and reinstates labor. Journal of Economic Perspectives, 33(2), 3-30. or are not shared with consumers. If that is the case, please document OS2 as “not present” when performing the Job Impact Assessment.

Caveat 3: Productivity boosts can significantly hamper job quality

Introduction of an AI system can lead to productivity enhancement through various routes: by allowing workers to produce more output per hour of work at the same level of effort or by allowing management to induce a higher level of effort from workers. If productivity boosts are expected to be achieved solely or mainly through increasing work intensity, please document OS2 as “not present” when performing the Job Impact Assessment.

Lastly, frontline workersBell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611 reported appreciation for AI systems that boosted their productivity by assisting them with core tasks. Conversely, technologies that boosted productivity by automating workers’ core tasks were associated with a reduction in job satisfaction.Valentine, M., and Hinds, R. (2022). How Algorithms Change Occupational Expertise by Prompting Explicit Articulation and Testing of Experts’ Theories. https://tinyurl.com/pxyr8ev3 Hence, pursuit of productivity increases through technologies that eliminate non-core tasks is preferred over paths that involve eliminating core tasks. Examples of technologies that assist workers on their core tasks include:

  • Training and coaching tools
  • Algorithmic decision support systems that give users additional information, analytics, or recommendations without prescribing or requiring decisions
OS3. Create new paid tasks for workers

Will the AI system create new tasks for humans or move unpaid tasks into paid work? Technological innovations have a great potential for benefit when they create new formal sector jobs, tasks, or markets that did not exist before. Consider, for example, the rise of social media influencers and content creators. These types of jobs were not possible before the rise of contemporary media and recommendation technologies. It has been estimated that, in 2018, more than 60 percent of employees were employed in occupations that did not exist in 1940.Autor, D. (2022). The labor market impacts of technological change: From unbridled enthusiasm to qualified optimism to vast uncertainty (No. w30074). National Bureau of Economic Research.

Caveat 1: Someone’s unpaid tasks can be someone else’s full-time job

 

It is important to keep in mind that technologies seemingly moving unpaid tasks into paid ones might, upon closer inspection, be producing an unintended (or deliberately unadvertised) effect of shifting tasks between paid jobs — often accompanied by a job quality downgrade. For example, a technology that allows people to hire someone to do their grocery shopping might convert their unpaid task into someone else’s paid one, but also reduce the demand for full-time domestic help workers, increasing precarity in the labor market.

Caveat 2: New tasks often go unacknowledged and unpaid

Sometimes the introduction of an AI system adds unacknowledged and uncompensated tasks to the scope of workers. For example, the labor of smoothing the effects of machine malfunction remains under the radar in many contexts,Mateescu, A., and Elish, M. (2019). AI in context: the labor of integrating new technologies. creating significant unacknowledged burdens on workers who end up responsible for correcting machine’s errors (without being adequately positioned to do that).Elish, M. C. (2019). Moral crumple zones: Cautionary tales in human-robot interaction (pre-print). Engaging Science, Technology, and Society (pre-print).

When performing the Job Impact Assessment, please explicitly document the applicability of these two caveats associated with OS3 for the AI system being assessed and its deployment context.

OS4. Support an egalitarian labor market
Will the AI system support a more egalitarian labor market structure? A superstar labor market structure is a situation where a relatively small number of workers dominate the market or satisfy most of the labor demand that exists in it. The opposite is an “egalitarian” labor structure where each worker’s output is small relative to the output of all other workers in the industry. The key factor that makes a labor market’s structure egalitarian is the presence of a need to invest an additional unit of worker time to serve an additional consumer. For example, the rise of the music recording industry has made its labor market structure less egalitarian for musicians. Today, to satisfy the demand for music from an additional customer, musicians do not need to physically get in front of them or do any additional work.
OS5. Be appropriate for lower-income geographies
Will the AI system be appropriate for lower-income geographies? Capital and labor of various skill types can be relatively more or less abundant in different countries. Technologies that take advantage of the factor of production (capital or labor of a certain skill type) that is relatively more abundant in a given country and do not require much of a factor that is relatively scarce there are deemed appropriate for that country.Generally, capital is relatively more abundant in the higher-income countries while labor is relatively more abundant in the lower-income countries, many of which also struggle with poor learning outcomes limiting the training the workforce receives.World Bank. (2017). World development report 2018: Learning to realize education’s promise. The World Bank. Therefore, capital-intensive labor-saving AI systems are generally inappropriate for lower-income countries whose main comparative advantage is relatively abundant labor.Korinek, A., and Stiglitz, J.E. (2021). Artificial intelligence, globalization, and strategies for economic development (No. w28453). National Bureau of Economic Research. Such technologies being adopted by high-income countries can hurt economic outcomes in lower-income countries because competitive forces in the export industries force the latter to adopt those technologies to remain competitive.Diao, X., Ellis, M., McMillan, M. S., and Rodrik, D. (2021). Africa’s manufacturing puzzle: Evidence from Tanzanian and Ethiopian firms (No. w28344). National Bureau of Economic Research.Rodrik, D. (2022). 4 Prospects for global economic convergence under new technologies. An inclusive future? Technology, new dynamics, and policy challenges, 65.

Consequently, lower-income countries would greatly benefit from access to technologies that would allow them to stay competitive by leveraging their abundant labor resources and creating gainful jobs that do not require high levels of educational attainment.

When assessing the presence of this signal, please also document if and how the relative abundance of capital and labor of various skill types is expected to change over time.

OS6. Broaden access to the labor market

Will the AI system broaden access to the labor market? AI systems that allow communities with limited or no access to formal employment to get access to gainful formal sector jobs are highly desirable from the perspective of broadly shared prosperity. Examples include AI systems that:

  • Assist the disabled
  • Make it easier to combine work and caregiving responsibilities
  • Enable work in languages the worker does not have a fluent command of
OS7. Boost revenue share of workers and society
Will the AI system boost workers’ and society’s share of an organization’s revenue? Workers’ share of revenue is the percentage of an organization’s revenue spent on workers’ wages and benefits. For the purposes of these Guidelines, we suggest excluding C-suite compensation when calculating workers’ share.If, following the introduction of an AI system, workers’ share of organization’s revenue is expected to grow or at least stay constant, it is a very strong signal that the AI system in question will serve to advance shared prosperity. The opposite is also true. If, following the introduction of an AI system, workers’ share of organization’s revenue is expected to shrink, it is a very strong signal that the AI system in question will harm shared prosperity.

Please note that worker benefits are included in workers’ share of an organization’s revenue. For example, consider an organization that adopts a productivity-enhancing AI system which allows it to produce the same or greater amount of output with fewer hours of work needed from human workers. That organization can decide to retain the same size of the workforce and share productivity gains with it (for example, in the form of higher wages, longer paid time off, or shorter work week at constant weekly pay), keeping the workers’ share of revenue constant or growing. That would be a prime example of using AI to advance shared prosperity.

Lastly, if an organization was able to generate windfall gains from AI development or usage and is committed to sharing the gains not only with workers it directly employs but the rest of the world’s population as well, that can be a great example of using AI to advance shared prosperity. While some have proposed this,O’Keefe, C., Cihon, P., Garfinkel, B., Flynn, C., Leung, J., and Dafoe, A. (2020, February). The windfall clause: Distributing the benefits of AI for the common good. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society (pp. 327-331). more research is needed to design mechanisms for making sure windfall gains are distributed equitably and organizations can be expected to reliably honor their commitment to distribute their gains.

OS8. Respond to needs expressed by impacted workers
Did workers who will use the AI system or be affected by it (or their representatives) identify the need for the system? AI systems created from a worker’s idea or identified need build in workers’ job expertise and preferences from the outset, making it more likely the AI systems will be beneficial or useful to workers affected by them and welcomed as such. Much of the current AI development pipeline starts with advances in research and development, only later identifying potential applications and product-market fit. The market for workplace AI technology is largely composed of senior executives and managers, creating a potential misalignment between needs perceived by budget holders and managers and the needs perceived by the workers who use or are most affected by the technology. AI systems emerging from the ideas and needs of workers who use or are most affected by them (or their representatives, who represent the collective voice of a given set of workers, not just the perspective of an individual worker) reduce this potential for misalignment.Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
OS9. Be co-developed with impacted workers

Were workers who will ultimately use or be affected by the AI system (or their representatives) included and given agency in every stage of the system’s development? Workers are subject matter experts in their own tasks and roles, and can illuminate opportunities and challenges for new technologies that are unlikely to be seen by those with less familiarity with the specifics of the work. The wisdom of workers who use or are most affected by AI systems introduced throughout development can smooth many rough edges that other contributors might only discover after systems are in the market and implemented. Where relevant worker representatives exist, they should be brought into the development process to represent collective worker interests from start to finish.Fully offering affected workers agency in the development process requires taking the time to understand their vantage points, and equip them or their representatives with enough knowledge about the proposed technology to meaningfully participate. They also must be afforded sufficient decision-making power to steer projects and, if necessary, end them in instances where unacceptable harms cannot be removed or mitigated. This also necessitates protecting their ability to offer suggestions freely without fear of repercussions. Without taking these steps, participatory processes can still lead to suboptimal outcomes — and possibly create additional harms through covering problems with a veneer of worker credibility.

OS10. Improve job quality or satisfaction
Was the AI system intended to improve job quality or increase job satisfaction? AI technology has the potential to improve many aspects of job quality and job satisfaction, from increasing occupational safety to providing personalized coaching that leads to career advancement. This requires taking job quality, worker needs, and worker satisfaction seriously. Two important caveats are required for this signal.

Caveat 1: Systems can improve one aspect of job quality while harming another

For example, many AI technologies positioned as safety enhancements are in reality invasive surveillance technologies. Though safety improvements may occur, harms to human rights, stress rates, privacy, job autonomy, job intensity, and other aspects of job quality may occur as well. Other AI systems purport to improve job quality by automating tasks workers dislike (see RS1 for more detail on the risks of task elimination).

When a system enhances one aspect of job quality while endangering another, this signal can still be counted as “present,” but the need to consider the rest of the opportunity and risk signals is particularly important.

Caveat 2: AI systems are sometimes deployed to redress job quality harms created by other AI systems

For example, some companies have introduced AI safety technologies to correct harms resulting from the prior introduction of an AI performance target-setting system that encouraged dangerous overwork.Scherer, M., and Brown, L. X. (2021). Warning: Bossware May Be Hazardous to Your Health. Center for Democracy and Technology. https://cdt.org/wp-content/uploads/2021/07/2021-07-29-Warning-Bossware-May-Be-Hazardous-To-Your-Health-Final.pdf

When this is the case, the introduction of the new AI system to redress the harms of the old does not count for this signal and should be marked as “not present.”

Instead of introducing new AI systems with their own attendant risks, the harms from the existing systems should be addressed in line with the Responsible Practices provided by the Guidelines for AI-using organizations and additional case-specific mitigations.

Signals of Risk to Shared Prosperity

Signals of Risk to Shared Prosperity

If one or more of the statements below apply to the AI system being assessed, this indicates a possibility of a negative impact on shared prosperity-relevant outcomes.

For-profit companies might feel pressure from investors to cut their labor costs no matter the societal price. We encourage investors and governments to join civil society in an effort to incentivize responsible business behavior with regards to shared prosperity and labor market impact.

Some practices or outcomes included in this section are illegal in some jurisdictions, and as such are already addressed in those locations. We include them here due to their legality in other jurisdictions.

Some of the signals of risk to shared prosperity described in the Guidelines are actively sought by companies as profit-making opportunities. The Guidelines do not suggest that companies should stop seeking profits, just that they should do so responsibly.

Profit-generating activities do not necessarily have to harm workers and communities, but some of them do. The presence of signals of risk indicates the potential to impose undue costs on society, which require mitigation strategies.

A risk signal (RS) is present if an AI system may:

Task-related Risks
RS1. Eliminate a given job’s core tasks
Will the AI system eliminate a significant share of tasks for a given job? A lot of technological innovations eliminate some job tasks that were previously done by human workers. That is not necessarily an unwelcome development, especially when those technologies also create new paid tasks for humans (see OS3), boost job quality (see OS10), or bring significant broadly distributed benefits (see OS1). For example, it can be highly desirable to automate tasks posing unmitigable risks to workers’ physical or mental health. Primary research conducted by the AI and Shared Prosperity Initiative indicated that frontline workers often experience automation of their non-core tasks as helpful and productivity-boosting.Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611

However, if an AI system is primarily geared towards eliminating core paid tasks without much being expected in terms of increased job quality or broadly shared benefits, nor in terms of new tasks for humans being created in parallel, then it warrants further attention as posing a risk to shared prosperity. The introduction of such a system will likely lower the demand for human labor, and thus wage or employment levels for affected workers.Acemoglu, D., and Restrepo, P. (2022). Tasks, automation, and the rise in US wage inequality. Econometrica, 90(5), 1973-2016. Automation of core tasks can also be experienced by workers as directly undermining their job satisfaction since workers’ core responsibilities are closely tied to their sense of pride and accomplishment in their jobs. For workers who see their jobs as an important part of their identity, core tasks are a major aspect of how they see themselves in the world.Valentine, M., and Hinds, R. (2022). How Algorithms Change Occupational Expertise by Prompting Explicit Articulation and Testing of Experts’ Theories. https://tinyurl.com/pxyr8ev3 Automation of core tasks can also lower the skill requirements of a job and reduce the formation of skills needed to advance to the next level.Nurski, L., and Hoffmann, M. (2022). The Impact of Artificial Intelligence on the Nature and Quality of Jobs. Working Paper. Bruegel. https://tinyurl.com/jxayzdcz

Please note that to evaluate the share of a given job’s tasks being eliminated, those tasks should be weighted by their importance for the production of the final output. We consider task elimination above 10% significant enough to warrant attention.

RS2. Reallocate tasks to lower-paid or more precarious jobs
Will the AI system enable reallocation of tasks to lower-paid or more precarious jobs or informal or unpaid labor? Often, while not eliminating human tasks on balance, AI technology enables shifting tasks from full-time jobs to unpaid or more precarious labor. The latter can happen, for example, through the “gig-ification” of work: technologically enabled separation of “time on task” and “idle time” which leads to unstable and unpredictable wages as well as the circumvention of minimum wage laws.

Paid tasks can also be converted into unpaid when new technology enables them to be performed by customers. Examples of that are self-checkout kiosks or automated customer support.Pritchett, L. (2020). The future of jobs is facing one, maybe two, of the biggest price distortions ever. Middle East Development Journal, 12(1), 131-156.

RS3. Reallocate tasks to higher- or lower-skilled jobs
Will the AI system enable the reallocation of tasks to jobs with higher or lower specialized skills requirements? Jobs with higher specialized skills requirements generally are better compensated, hence an AI system shifting tasks into such jobs will likely lead to a positive effect of more of them being opened up. However, those jobs might not be accessible to people affected by task reallocation because those people might not possess the newly required specialized skills. Retraining and job matching support programs can help here, though those often fall short. Word processor is an example of a technology that reallocated typing-related tasks away from typists to managers. Generative AI applications are an example of a recent technology anticipated to induce broad-reaching shifts in skill requirements of large swaths of jobs.Eloundou, T., Manning, S., Mishkin, P., and Rock, D. (2023). GPTs are GPTs: An early look at the labor market impact potential of large language models. arXiv preprint arXiv:2303.10130.Noy, S., and Zhang, W. (2023). Experimental evidence on the productivity effects of generative artificial intelligence. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4375283Korinek, A. (2023). Language models and cognitive automation for economic research (No. w30957). National Bureau of Economic Research.

Importantly, AI-induced reallocation of tasks to jobs with lower specialized skills requirements may be positive but is still a risk signal warranting further attention, because lowering specialized skill requirements can lower not only the barriers to entry to the occupation, but also prevailing wages.

Market-related Risks
RS4. Move jobs away from geographies with few opportunities
Will the AI system move job opportunities away from geographies where there would be few remaining? Due to associated costs and excessive immigration barriers, labor mobility remains low, both within and between countries. As a result, changes that move job opportunities from one area to another can harm workers in the losing area. Research suggests that disappearance of stable, well-paying jobs can profoundly re-shape regions, leading to a rise in “deaths of despair,” addictions, and mental health problems.Case, A., and Deaton, A. (2020). Deaths of Despair and the Future of Capitalism. Princeton University Press.Gihleb, R., Giuntella, O., Stella, L., and Wang, T. (2022). Industrial robots, workers’ safety, and health. Labour Economics, 78, 102205. Impacted communities might be able to bounce back from job loss if comparable alternative job opportunities are sufficiently available in their area. But even when those exist, the presence of labor market frictions make it important to invest in creating support programs to help workers move into new jobs of comparable quality.

In addition to jobs disappearing as the direct effect of labor-saving technology being introduced in a region, please note that this effect can also be an indirect result of labor-saving technology initially introduced in a completely different region or country. Due to excessive immigration barriers, AI developers based in high-income countries face massively inflated incentives to create labor-saving technologies far in excess of what would be socially optimal given the world’s overall level of labor supply/demand for jobs.Pritchett, L. (2020). The future of jobs is facing one, maybe two, of the biggest price distortions ever. Middle East Development Journal, 12(1), 131-156. Once that technology is developed in the high-income countries it gets deployed all over the world, including countries facing a dire need of formal sector jobs.Pritchett, L. (2023). Choose People. LaMP Forum. https://lampforum.org/2023/03/02/choose-people/

RS5. Increase market concentration and barriers to entry
Will an AI system increase market concentration and barriers to market entry? An increase in market concentration is a signal of a possible labor market impact to come for at least two reasons:

  • It increases the risk of job cuts by competing firms
  • It makes it less likely that the winning firm shares efficiency gains with workers in the form of better wages/benefits or with consumers in the form of lower prices/higher-quality products

Therefore, in a monopolistic market, any benefits brought on by AI are likely to be shared by few, while the harms might still be widely distributed. Similarly, job impacts that might occur in upstream or downstream industries due to an AI-induced increase in market concentration need to be accounted for as well.

Sourcing-related Risks
RS6. Rely on poorly treated or compensated outsourced labor
Will the AI system rely on, for either model training or operation, outsourced labor deprived of a living wage and decent working conditions? The process of building datasets for model training can be highly labor-intensive. It often requires human workers (whom we will refer to as data enrichment professionals) to review, classify, annotate, and otherwise manage massive amounts of data. Despite the foundational role played by data enrichment professionals, a growing body of research reveals the precarious working conditions that they face, which include:Gray, M. L., and Suri, S. (2019). Ghost work: How to stop Silicon Valley from building a new global underclass. Eamon Dolan Books.

  • Inconsistent and unpredictable compensation for their work
  • Unfairly rejected and therefore unpaid labeling tasks
  • Long, ad-hoc working hours
  • Lack of means to contest or get an explanation for the decisions affecting their take-home pay and ratings

Lack of transparency around data enrichment labor sourcing practices in the AI industry exacerbate this issue.

RS7. Use training data collected without consent or compensation
Will the AI system be trained using a dataset containing data collected without consent and/or compensation? AI systems can be trained on data that embeds the economically-relevant know-how of people who generated that data, which can be especially problematic if the subsequent deployment of that AI system reduces the demand for labor of those people. Examples include but are not limited to:

  • Images created by artists and photographers that are used to train generative AI systems
  • Keystrokes and audio recordings of human customer service agents used to create automated customer service routines
  • Records of actions taken by human drivers used to train autonomous driving systems
Worker Abuse-related Risks
RS8. Predict the lowest wages a worker will accept
Will the AI system be used to predict the lowest wage a given worker would accept? It has been documented that workers can experience the impact of AI systems used for workforce management as effectively depriving them of being able to predict their take-home wages with any amount of certainty.Dubal, V. (2023). On Algorithmic Wage Discrimination. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4331080 An AI system allowing predictions about the lowest wages an individual worker would accept is analogous to a system allowing for perfect price discrimination of consumers. Price discrimination, while always driven by monopoly power and thus inefficient, is considered acceptable in certain situations, such as reduced price of museum admission for seniors and students. However, that acceptability is predicated on the transparency of the underlying logic. A possibility of using an algorithmic system to create take-home pay “personalization,” especially based on logic that is opaque to the workers or ever-changing, should serve as a strong signal of a potential negative impact on shared prosperity. A related risk for informal workers is the use of AI to reduce their bargaining power relative to those they contract with. Information asymmetries created through AI use by purchasers of their work are an emerging risk to workers in the informal sector.Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
RS9. Accelerate task completions without other changes
Will the AI system accelerate task completion without meaningfully changing resources, tools, or skills needed to accomplish the tasks? Some AI systems push workers to higher performance on goals, targets, or KPIs without modifying how the work is done. Examples of this include speeding up the pace with which workers are expected to complete tasks or using AI to set performance goals that are just out of reach for many workers. When this occurs without additional support for workers in the form of streamlining, simplifying, or otherwise improving the process of completing the task, it risks higher stress and injury rates for workers.
RS10. Reduce schedule predictability
Will the AI system reduce ​​the amount of advance notice a worker receives regarding changes to their working hours? Schedule predictability is strongly tied to workers’ physical and mental health.Schneider, D., and Harknett, K. (2017, April). Schedule Instability and Unpredictability and Worker and Family Health and Well-being. In PAA 2017 Annual Meeting. PAA.Williams, J. et al. (2022). Stable scheduling study: Health outcomes report. https://ssrn.com/abstract=4019693 Automated, last-minute scheduling software can harm workers’:

  • Emotional well-being through increased stress
  • Occupational safety and health through sleep deprivation/unpredictability and the physical effects of stress
  • Financial well-being through missed shifts and increased need for more expensive transit (for example, ride-hailing services at times when public transit isn’t frequent or safe).

Recent AI technology designed to lower labor costs by reducing the number of people working during predicted “slow” times has disrupted schedule predictability, with workers receiving minimal notice about hours that have been eliminated from or added to their schedules.

RS11. Reduce workers’ break time
Will the AI system infringe on workers’ breaks or encourage them to do so? Workers’ breaks are necessary for their recovery from physically, emotionally, or intellectually strenuous or intense periods of work, and are often protected by law. Some AI systems billed as productivity software infringe on workers’ breaks by sending them warnings based on the time they’ve spent away from their workstations or “off-task,” even during designated breaks or while they are using allotted break time.Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611 Others implicitly encourage workers to skip breaks by setting overly ambitious performance targets that pressure workers to work through downtime to meet goals. These systems can foster higher rates of injury or stress, undermine focus, and reduce opportunities to form social relationships at work.
RS12. Increase overall difficulty of tasks
Will the AI system increase the overall difficulty of tasks? When AI systems are used to automate less demanding tasks (for example, the most straightforward, emotionally neutral customer requests in a call center), workers may be left with a higher concentration of more demanding tasks, effectively increasing the difficulty of their job.Dzieza, J. (2020). Robots aren’t taking our jobs — They’re becoming our bosses. The Verge. https://tinyurl.com/5a9mxeuz Difficulty increases may take the form of more physically, emotionally, or intellectually demanding tasks. The higher intensity may also place them at higher risk of burning out. While some workers may welcome the added challenge, the above concerns merit caution, especially if workers are not compensated equitably for the increased difficulty.
RS13. Enable detailed monitoring of workers
Will the AI system monitor something other than the pace and quality of task completion? The use of AI to monitor workers is just the latest entry in the long history of the technological surveillance of labor.Levy, K. (2022). Data Driven: truckers, technology, and the new workplace surveillance. Princeton University Press. However, AI capabilities have increased the frequency, comprehensiveness, and intensiveness of on-the-job monitoring. This use of AI often extends beyond monitoring of workers’ direct responsibilities and outputs, including information as varied as their time in front of their computer or time spent actively using their computer, their movements through an in-person worksite, and the frequency and content of communications with other workers. This detailed monitoring risks:

  • Increasing stress and anxiety
  • Harming their privacy
  • Causing them to feel a lack of trust from their employer
  • Undermining their sense of autonomy on the job
  • Lowering engagement and job satisfaction
  • Chilling worker organizing, undermining worker voice.Moore, P.V. (2017). The quantified self in precarity: Work, technology and what counts. Routledge.Scherer, M., and Brown, L. X. (2021). Warning: Bossware May Be Hazardous to Your Health. Center for Democracy and Technology. https://cdt.org/wp-content/uploads/2021/07/2021-07-29-Warning-Bossware-May-Be-Hazardous-To-Your-Health-Final.pdf.

While monitoring systems can have legitimate uses (such as enhancing worker safety), even good systems can be abused, particularly in environments with low worker agency or an absence of regulations, monitoring, and enforcement of worker protections.Brand, J., Dencik, L. and Murphy, S. (2023). The Datafied Workplace and Trade Unions in the UK. Data Justice Lab. https://datajusticeproject.net/wp-content/uploads/sites/30/2023/04/Unions-Report_final.pdf.

RS14. Reduce worker autonomy
Will the AI system reduce workers’ autonomy, decision-making authority, or control over how they complete their work? Autonomy, decision-making authority, job control, and the exercise of discernment in performing one’s job are correlated with high job quality and job satisfaction.Nurski, L., and Hoffmann, M. (2022). The Impact of Artificial Intelligence on the Nature and Quality of Jobs. Working Paper. Bruegel. https://tinyurl.com/2a943p8f Reducing scope for these activities could also be a sign of a shift from a “high-road” staffing approach (where experience and expertise is valued) to a “low-road” approach (where less training or experience is needed and thus workers hold less bargaining power and can be more easily replaced). In the informal sector, this may appear as a reduction in the scope for design and creativity by artisans and garment workers.Nanavaty, R. (2023). Interview with Reema Nanavaty, Self-Employed Women’s Association.
RS15. Reduce mentorship or apprenticeship opportunities
Will the AI system reduce workers’ opportunities for mentorship or apprenticeship? Automated training, automated coaching, and automation of entry-level tasks may lower workers’ opportunities for apprenticeship and mentorship. Apprenticeship is an important way for workers to learn on the job, and develop the skills they need to advance.Beane, M. (2022). Today’s Robotic Surgery Turns Surgical Trainees into Spectators: Medical Training in the Robotics Age Leaves Tomorrow’s Surgeons Short on Skills. IEEE Spectrum, 59(8), 32-37. https://tinyurl.com/wyhxukhk Mentorship and apprenticeship can help workers develop social relationships and community with peers and supervisors. Additionally, mentors can help workers learn to navigate unspoken rules and norms in the workplace, and assist them with career development within and beyond their current workplace.
RS16. Reduce worker satisfaction
Will the AI system reduce the motivation, engagement, or satisfaction of the workers who use it or are affected by it? While this test directly speaks to meaning, community, and purpose, it is also a proxy for other aspects of worker well-being. Demotivation and disengagement are signs of lowered job satisfaction and serve as indications of other job quality issues.
RS17. Influence employment and pay decisions
Will the AI system make or suggest decisions on recruitment, hiring, promotion, performance evaluation, pay, wage penalties, and bonuses? The decisions outlined in this signal are deeply meaningful to workers, meriting heightened attention from employers. Automation of these decisions should raise concern, as automated systems might lack the complete context necessary for these decisions and risk subjecting workers to “algorithmic cruelty.”Gray, M. L., and Suri, S. (2019). Ghost work: How to stop Silicon Valley from building a new global underclass. Eamon Dolan Books. They also risk introducing additional discriminatory bases for decisions, beyond those already existent in human decisions.Center for Democracy and Technology et al. 2022 In instances where AI systems are used to suggest (rather than decide) on these questions, careful implementation focused on increasing decision accuracy and transparency can benefit workers. However, human managers using these systems often find it undesirable or difficult to challenge or override recommendations from AI, making the system’s suggestions more binding than they may initially appear and meriting additional caution in these uses.
RS18. Operate in discriminatory ways
Will the AI system operate in ways that are discriminatory? AI systems have been repeatedly shown to reproduce or intensify human discrimination patterns on demographic categories such as gender, race, age, and more.Buolamwini, J., and Gebru, T. (2018, January). Gender shades: Intersectional accuracy disparities in commercial gender classification. In Conference on fairness, accountability and transparency (pp. 77-91). PMLR.Benjamin, R. (2019). Race After Technology: Abolitionist Tools for the New Jim Code. John Wiley and Sons.Keyes, O. (2018). The misgendering machines: Trans/HCI implications of automatic gender recognition. Proceedings of the ACM on human-computer interaction, 2(CSCW), 1-22.Rosales, A., and Fernández-Ardèvol, M. (2019). Structural ageism in big data approaches. Nordicom Review, 40(s1), 51-64. Workplace AI systems should be rigorously tested to ensure that they operate fairly and equitably.


AI-Creating Organizations ➜


AI-Using Organizations ➜


Policymakers ➜


Labor Organizations and Workers ➜

Guidelines for AI and Shared Prosperity

Home

Step 1: Learn About the Guidelines

The Need for the Guidelines

The Origin of the Guidelines

Design of the Guidelines

Key Principles for Using the Guidelines

Step 2: Apply the Job Impact Assessment Tool

Instructions for Performing a Job Impact Assessment

Signals of Opportunity to Advance Shared Prosperity

Signals of Risk to Shared Prosperity

STEP 3: Stakeholder-Specific Recommendations

For AI-Creating Organizations

For AI-Using Organizations

For Policymakers

For Labor Organizations and Workers

Get Involved

Endorsements

Acknowledgments

AI and Shared Prosperity Initiative’s Steering Committee

Sources Cited

  1. ​​Acemoglu, D. (Ed.). (2021). Redesigning AI: Work, democracy, and justice in the age of automation. Boston Review.
  2. Korinek, A., and Stiglitz, J.E. (2020, April). Steering technological progress. In NBER Conference on the Economics of AI.
  3. Acemoglu, D., and Johnson, S. (2023). Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. Public Affairs, New York.
  4. International Labour Organization. (n.d.). Decent work. https://tinyurl.com/yur776yd
  5. US Department of Commerce and US Department of Labor. (n.d.). Department of Commerce and Department of Labor Good Jobs Principles, DOL. https://tinyurl.com/mtbpemkn
  6. Institute for the Future of Work. (n.d.). The Good Work Charter. https://tinyurl.com/ycxtaax4
  7. Klinova, K., and Korinek, A. (2021). AI and shared prosperity. In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society (pp. 645-651).
  8. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  9. Partnership on AI, 2021. Redesigning AI for Shared Prosperity: an Agenda. https://partnershiponai.org/paper/redesigning-ai-agenda/
  10. Negrón, W. (2021). Little Tech is Coming for Workers. Coworker.org. https://home.coworker.org/wp-content/uploads/2021/11/Little-Tech-Is-Coming-for-Workers.pdf.
  11. Korinek, A., 2022. How innovation affects labor markets: An impact assessment.
  12. Brynjolfsson, E., Collis, A., Diewert, W.E., Eggers, F., and Fox, K.J. (2019). GDP-B: Accounting for the value of new and free goods in the digital economy (No. w25695). National Bureau of Economic Research.
  13. Acemoglu, D., and Restrepo, P. (2019). Automation and new tasks: How technology displaces and reinstates labor. Journal of Economic Perspectives, 33(2), 3-30.
  14. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  15. Valentine, M., and Hinds, R. (2022). How Algorithms Change Occupational Expertise by Prompting Explicit Articulation and Testing of Experts’ Theories. https://tinyurl.com/pxyr8ev3
  16. Autor, D. (2022). The labor market impacts of technological change: From unbridled enthusiasm to qualified optimism to vast uncertainty (No. w30074). National Bureau of Economic Research.
  17. Mateescu, A., and Elish, M. (2019). AI in context: the labor of integrating new technologies.
  18. Elish, M. C. (2019). Moral crumple zones: Cautionary tales in human-robot interaction (pre-print). Engaging Science, Technology, and Society (pre-print).
  19. World Bank. (2017). World development report 2018: Learning to realize education's promise. The World Bank.
  20. Korinek, A., and Stiglitz, J.E. (2021). Artificial intelligence, globalization, and strategies for economic development (No. w28453). National Bureau of Economic Research.
  21. Diao, X., Ellis, M., McMillan, M. S., and Rodrik, D. (2021). Africa's manufacturing puzzle: Evidence from Tanzanian and Ethiopian firms (No. w28344). National Bureau of Economic Research.
  22. Rodrik, D. (2022). 4 Prospects for global economic convergence under new technologies. An inclusive future? Technology, new dynamics, and policy challenges, 65.
  23. O'Keefe, C., Cihon, P., Garfinkel, B., Flynn, C., Leung, J., and Dafoe, A. (2020, February). The windfall clause: Distributing the benefits of AI for the common good. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society (pp. 327-331).
  24. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  25. Scherer, M., and Brown, L. X. (2021). Warning: Bossware May Be Hazardous to Your Health. Center for Democracy and Technology. https://cdt.org/wp-content/uploads/2021/07/2021-07-29-Warning-Bossware-May-Be-Hazardous-To-Your-Health-Final.pdf
  26. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  27. Acemoglu, D., and Restrepo, P. (2022). Tasks, automation, and the rise in US wage inequality. Econometrica, 90(5), 1973-2016.
  28. Valentine, M., and Hinds, R. (2022). How Algorithms Change Occupational Expertise by Prompting Explicit Articulation and Testing of Experts’ Theories. https://tinyurl.com/pxyr8ev3
  29. Nurski, L., and Hoffmann, M. (2022). The Impact of Artificial Intelligence on the Nature and Quality of Jobs. Working Paper. Bruegel. https://tinyurl.com/jxayzdcz
  30. Pritchett, L. (2020). The future of jobs is facing one, maybe two, of the biggest price distortions ever. Middle East Development Journal, 12(1), 131-156.
  31. Eloundou, T., Manning, S., Mishkin, P., and Rock, D. (2023). GPTs are GPTs: An early look at the labor market impact potential of large language models. arXiv preprint arXiv:2303.10130.
  32. Noy, S., and Zhang, W. (2023). Experimental evidence on the productivity effects of generative artificial intelligence. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4375283
  33. Korinek, A. (2023). Language models and cognitive automation for economic research (No. w30957). National Bureau of Economic Research.
  34. Case, A., and Deaton, A. (2020). Deaths of Despair and the Future of Capitalism. Princeton University Press.
  35. Gihleb, R., Giuntella, O., Stella, L., and Wang, T. (2022). Industrial robots, workers’ safety, and health. Labour Economics, 78, 102205.
  36. Pritchett, L. (2020). The future of jobs is facing one, maybe two, of the biggest price distortions ever. Middle East Development Journal, 12(1), 131-156.
  37. Pritchett, L. (2023). Choose People. LaMP Forum. https://lampforum.org/2023/03/02/choose-people/
  38. Gray, M. L., and Suri, S. (2019). Ghost work: How to stop Silicon Valley from building a new global underclass. Eamon Dolan Books.
  39. Dubal, V. (2023). On Algorithmic Wage Discrimination. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4331080
  40. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  41. Schneider, D., and Harknett, K. (2017, April). Schedule Instability and Unpredictability and Worker and Family Health and Well-being. In PAA 2017 Annual Meeting. PAA.
  42. Williams, J. et al. (2022). Stable scheduling study: Health outcomes report. https://ssrn.com/abstract=4019693
  43. Bell, S. A. (2022). AI and Job Quality: Insights from Frontline Workers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4337611
  44. Dzieza, J. (2020). Robots aren’t taking our jobs — They’re becoming our bosses. The Verge. https://tinyurl.com/5a9mxeuz
  45. Levy, K. (2022). Data Driven: truckers, technology, and the new workplace surveillance. Princeton University Press.
  46. Moore, P.V. (2017). The quantified self in precarity: Work, technology and what counts. Routledge.
  47. Scherer, M., and Brown, L. X. (2021). Warning: Bossware May Be Hazardous to Your Health. Center for Democracy and Technology. https://cdt.org/wp-content/uploads/2021/07/2021-07-29-Warning-Bossware-May-Be-Hazardous-To-Your-Health-Final.pdf.
  48. Brand, J., Dencik, L. and Murphy, S. (2023). The Datafied Workplace and Trade Unions in the UK. Data Justice Lab. https://datajusticeproject.net/wp-content/uploads/sites/30/2023/04/Unions-Report_final.pdf.
  49. Nurski, L., and Hoffmann, M. (2022). The Impact of Artificial Intelligence on the Nature and Quality of Jobs. Working Paper. Bruegel. https://tinyurl.com/2a943p8f
  50. Nanavaty, R. (2023). Interview with Reema Nanavaty, Self-Employed Women’s Association.
  51. Beane, M. (2022). Today's Robotic Surgery Turns Surgical Trainees into Spectators: Medical Training in the Robotics Age Leaves Tomorrow's Surgeons Short on Skills. IEEE Spectrum, 59(8), 32-37. https://tinyurl.com/wyhxukhk
  52. Gray, M. L., and Suri, S. (2019). Ghost work: How to stop Silicon Valley from building a new global underclass. Eamon Dolan Books.
  53. Center for Democracy and Technology et al. 2022
  54. Buolamwini, J., and Gebru, T. (2018, January). Gender shades: Intersectional accuracy disparities in commercial gender classification. In Conference on fairness, accountability and transparency (pp. 77-91). PMLR.
  55. Benjamin, R. (2019). Race After Technology: Abolitionist Tools for the New Jim Code. John Wiley and Sons.
  56. Keyes, O. (2018). The misgendering machines: Trans/HCI implications of automatic gender recognition. Proceedings of the ACM on human-computer interaction, 2(CSCW), 1-22.
  57. Rosales, A., and Fernández-Ardèvol, M. (2019). Structural ageism in big data approaches. Nordicom Review, 40(s1), 51-64.
  58. Klinova, K. (2022) Governing AI to Advance Shared Prosperity. In Justin B. Bullock et al. (Eds.), The Oxford Handbook of AI Governance. Oxford Handbooks.
  59. Park, H., Ahn, D., Hosanagar, K., and Lee, J. (2021, May). Human-AI interaction in human resource management: Understanding why employees resist algorithmic evaluation at workplaces and how to mitigate burdens. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-15).
  60. Bernhardt, A., Suleiman, R., and Kresge, L. (2021). Data and algorithms at work: the case for worker technology rights. https://laborcenter.berkeley.edu/wp-content/uploads/2021/11/Data-and-Algorithms-at-Work.pdf.
  61. Colclough, C.J. (2022). Righting the Wrong: Putting Workers’ Data Rights Firmly on the Table. https://tinyurl.com/26ycnpv2
  62. Pasquale, F. (2020). New Laws of Robotics. Harvard University Press.
  63. Rodrik, D. (2022). 4 Prospects for global economic convergence under new technologies. An inclusive future? Technology, new dynamics, and policy challenges, 65.
  64. Anderson, E. (2019). Private Government: How Employers Rule Our Lives (and Why We Don’t Talk about it). Princeton University Press.
  65. Korinek, A. (2022). How innovation affects labor markets: An impact assessment.
  66. Institute for the Future of Work. (2023). Good Work Algorithmic Impact Assessment Version 1: An approach for worker involvement. https://tinyurl.com/mr4yn5yt
  67. Bernhardt, A., Suleiman, R., and Kresge, L. (2021). Data and algorithms at work: the case for worker technology rights. https://laborcenter.berkeley.edu/wp-content/uploads/2021/11/Data-and-Algorithms-at-Work.pdf.
  68. Colclough, C.J. (2022). Righting the Wrong: Putting Workers’ Data Rights Firmly on the Table. https://tinyurl.com/26ycnpv2
  69. Brand, J., Dencik, L. and Murphy, S. (2023). The Datafied Workplace and Trade Unions in the UK. Data Justice Lab. https://datajusticeproject.net/wp-content/uploads/sites/30/2023/04/Unions-Report_final.pdf.
  70. Park, H., Ahn, D., Hosanagar, K., and Lee, J. (2021, May). Human-AI interaction in human resource management: Understanding why employees resist algorithmic evaluation at workplaces and how to mitigate burdens. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-15).
  71. Mateescu, A., and Elish, M. (2019). AI in context: the labor of integrating new technologies.
  72. Elish, M. C. (2019). Moral crumple zones: Cautionary tales in human-robot interaction (pre-print). Engaging Science, Technology, and Society (pre-print).
Table of Contents