Ongoing discussions among technology industry leaders about the governance of fast-moving artificial intelligence have led some observers to argue that workplace-level AI challenges are receiving insufficient attention from business owners. Multiple polls revealed that a notable proportion of workers were deploying automation tools without adequate oversight, frequently in workplaces where no formal policies governed their responsible use.
Four independent research papers have mapped the overlapping obstacles that an expanding workforce and their organisations encounter as they navigate the opening stages of AI integration. A GoTo poll of 2,500 workers and technology professionals spanning several nations revealed that four in five respondents harboured worries about being held accountable — or losing their jobs — for mistakes in AI-generated workplace documents. Approximately one-third of those respondents reported holding back from adopting the technology, opting instead to complete the work on their own.
According to a Gallup survey, roughly one in four employed Americans expressed concern that automation or technological advancement could render their roles redundant. Since 2021, anxiety about job security among workers under 45 has climbed steadily, placing it at the top of employment-related concerns for that age group. This trend is also driving many of those workers to adopt AI tools as a means of boosting output and safeguarding their positions.
The GoTo survey measured this response, finding it was partly attributable to workplace pressure on employees to adopt AI tools. According to the findings, anxieties around workplace performance and organisational pressure to adopt AI tools are contributing to their improper use. Six in ten workers report experiencing pressure to adopt AI tools as a way to boost their output.
According to the poll, 47 percent of respondents reported that AI usage factors into their performance evaluations, while 42 percent indicated they would accept the risk of errors from AI tools rather than be perceived as inefficient or expendable. Around one in three participants noted that an unspoken expectation exists in their workplace to rely on AI without raising concerns about its inaccuracies. A separate survey found that one in seven workers reported being instructed to avoid alerting supervisors to errors in the application's output.
According to the GoTo survey, 55 percent of respondents reported that they do not consistently verify or validate AI-generated outputs, including for tasks with meaningful consequences. A quarter of those surveyed acknowledged that the technology enabled them to sidestep responsibility at work, with a similar proportion reporting they had attributed their own errors to the applications involved.
According to a September survey conducted by California-based litigation firm Kolmogorov Law, P.C., roughly 42 percent of the 50 AI-using workers polled reported having deliberately forwarded AI-generated content they believed or suspected to be inaccurate into broader organisational processes. Roughly three in ten of those employees reported facing adverse repercussions as a result. Nonetheless, nearly two-thirds of survey respondents indicated they had yet to check AI-generated content for factual accuracy.
According to the Kolmogorov survey, just 22 percent of participants indicated their organisation had a written policy mandating verification of AI-generated content before inclusion in work product, while 59 percent said no such formal requirement existed at their workplace. According to the GoTo survey, only about a quarter of small businesses and roughly a third of midsize companies have put AI policies in place. That proportion climbs to 53 percent among larger organisations.
Kolmogorov Law founder Pavel Kolmogorov noted in the report that employees appear to have recognised something many employers have yet to formalise in writing, and that verification remains largely absent. Citing an AI-generated response does not constitute an adequate defence once that output appears in a contract, a legal filing, or work delivered to a client. Warranty disclaimers built into AI companies' terms of service leave those who suffer harm without a clear legal target to hold accountable. Responsibility, therefore, falls on the professional and the business — precisely the domains where such obligations have long been recognised.
Both research efforts determined that establishing transparent and well-defined usage guidelines is the primary method through which organisations can boost employee output and ensure accurate application results. Dan Schawbel, managing partner of Workplace Intelligence, which carried out the GoTo survey, noted that workers currently face a dual pressure — feeling compelled to demonstrate AI use to remain relevant, yet anxious about accountability when errors occur — a dynamic he described as a pathway to careless AI practices across the workforce.
Organisations are advised to limit top-down mandates around AI adoption and instead pursue constructive approaches that account for the additional time employees require to thoroughly check tool outputs. A number of organisations are rolling out AI tools at pace, yet many have not put in place the guidance, workforce preparation, or operational frameworks that staff would need to use them effectively. Organisations that narrow this divide may find themselves better placed to harness artificial intelligence effectively while fostering a more capable and assured workforce.
