BCG's AI Ethics Officer says: AI's real value starts once employees…

Boston Consulting Group's Global Chief AI Ethics Officer Steven Mills says the key to unlocking AI's workplace potential lies in hands-on employee training—something most companies are failing to provide. Speaking at Semafor's World Economic Summit, Mills told Business Insider that employees want about five hours of hands-on training, coaching, and mentoring. However, only about a third of workers are actually receiving that level of instruction.The real transformation begins once employees experience AI's practical benefits firsthand, Mills explained. When workers use AI for simple tasks like editing email bullet points and see positive results, they immediately start thinking about broader applications, creating what he calls a "virtuous cycle."This training gap may help explain why BCG's recent research shows only 5% of companies are deriving meaningful value from AI investments. The September report, based on data from over 1,250 global firms, found that 60% of companies have seen minimal increases in revenue and cost savings despite substantial AI investments, Business Insider reported.Mills emphasised that organisations need to "reimagine the art of the possible" rather than treating AI as just another tool. "A big thing that organizations are not doing is stepping back and saying, 'How do we really reimagine our business processes, our service offerings, now that we have AI?'" he said.Mills, who also leads BCG's Center for Digital Government, noted that public sector AI adoption is accelerating faster than expected. Major AI companies including OpenAI, Anthropic, Meta, Google, and Microsoft have offered their AI agents to federal agencies at minimal cost."I think you'll see a big hockey stick in terms of rate of adoption here soon," Mills predicted, adding that people who use AI technology in their private lives naturally want access at work as well.The message is clear: companies that invest in proper employee training and reimagine their processes—rather than simply plugging AI into existing workflows—will be the ones capturing real value from their AI investments.