Knowledge Workers' Perspectives on AI Training for Responsible AI Use

Most knowledge workers lack adequate AI training despite widespread workplace adoption. The study found that 39 workers across HR, labor law, standards, and training roles from 26 countries identified significant gaps in how organizations prepare employees to use AI safely, raising concerns that inadequate training could lead workers to either abandon AI tools or misuse them in ways that perpetuate bias.

Workers struggle with fundamental AI literacy, particularly understanding AI limitations and misinterpreting results. Participants identified "understanding what AI is" and "misinterpreting outcomes" as critical training gaps, suggesting that even knowledge workers in specialized fields may not grasp how AI systems work or recognize when AI-generated results are unreliable or inappropriate for decision-making.

AI training must address bias amplification and worker rights concerns. The research surfaced nine key training topics, including how AI can exacerbate existing biases and how workers' rights may be affected by AI systems—issues that go beyond technical skills to encompass ethical and legal dimensions of responsible AI deployment.

Current training approaches risk perpetuating harmful values if not carefully designed. The researchers emphasized that training programs themselves must be scrutinized to ensure they don't reinforce problematic assumptions or normalize irresponsible AI use, suggesting that how organizations teach about AI matters as much as what they teach.

Context-specific training solutions are needed across different industries and regions. With participants spanning multiple countries and professional domains, the study highlights that effective AI training cannot be one-size-fits-all but must account for different workplace contexts, regulatory environments, and organizational needs.