Generative AI and the workforce: 10 big trends we're seeing right now

Those quickest to adopt GenAI among their workforce can be described as data-driven, the report says. These organizations have a long history of establishing robust data quality, infrastructure, governance and security. They may not necessarily be faster at identifying GenAI use cases, but when they do they already have everything in place to deploy quickly.

Many early adopters have moved beyond the initial experimentation phase. Among the key lessons they report is the importance of not rushing the implementation process and the benefits of testing solutions in small groups before a wider roll-out. This helps identify issues early and prevents employees losing interest if things don’t work as planned.

Most organizations featured in the report are highly conscious of the risks around deploying GenAI in the workforce. These include data breaches, privacy violations and bias in outcomes or other ethical aspects. To prevent reputational damage and avoid conflict with regulators and authorities, many are taking a cautious approach by conducting experiments and implementing pilots within the comparatively safe environment of their organization.

While the report says it is currently difficult to assess GenAI productivity gains at a macroeconomic level, at an organizational level, such gains are being reported. One company claimed requests that would have once taken weeks to complete now take just minutes with automation – an example of how gains are particularly evident in routine and repetitive work. Empowering people in this way is frequently mentioned in the report – more than a quarter of respondents said that GenAI enables employees to do more enjoyable, creative and value-adding work. However, a number of organizations quizzed do not have a clear plan for what workers should do with any freed-up time.

Productivity improvement is not the only driver for GenAI deployment, with improving the quality of work mentioned by respondents as equally important. If correctly implemented, the technology has the potential to be more accurate and consistent and make fewer mistakes than humans and can therefore lead to higher quality and customer satisfaction.

From concerns around accuracy to the potential existence of bias and the ethics of replacing human work with that of GenAI, employees have many questions about the use of the technology. Frequently, IT professionals within an organization were among the quickest to embrace GenAI. Meanwhile, the most significant impact is being felt in departments that conduct a lot of administrative work, leading to uncertainty among those teams. Trust can be built through training that demystifies the technology and reskilling and upskilling that gives workers the potential to grow into new roles. This will be vital in the near future – as 44% of workers’ skills will be disrupted in the next five years, according to the Forum’s Jobs Initiative, which is working towards good jobs for all in the context of such labour market disruptions.

With new initiatives, it is important to also understand the effect on the culture of the organization and the mindset shift it requires of employees, the report says. Effective leadership, from the very top of the organization, is vital. And there is a crucial role for middle managers, who understand workflows and processes and therefore where GenAI can have the biggest impact.

In fact, the interviewed companies reported varying figures from 20% to 80%. Some stated that almost everyone was using the technology, or at least that they could because the whole organization had been given access to GenAI tools. How accessible these tools are to workforces depends on a company’s appetite for risk – some respondents grant all employees access, while others limit to certain departments or require licences to be requested.

Compared to smaller, task-specific AI models, large language models such as ChatGPT are energy intensive, with each prompt requiring calculations that consume a significant amount of power. While this is a problem most organizations in the report acknowledge, few have yet developed a strategy for acting on it and environmental considerations do not seem to be central to GenAI workforce deployment decisions.

Most organizations interviewed for the report monitor the risks, quality and responsible use of GenAI through internal committees or councils, which establish rules and frameworks and assess use cases. Nearly all also say they have developed training programmes for the responsible use of tools. With knowledge of scandals around discriminatory algorithms and incoming legislation such as the European Union AI Act, companies are acutely aware of the importance of validation, verification and human intervention. “The biggest mistake you can make is to remove humans from your processes,” one report interviewee said.