IBM study finds 60 percent of workers fear AI is eroding skills

IBM’s study is a reminder that the binding constraint on AI deployment is quickly becoming human capability and governance, not just model quality. If most employees feel their skills are eroding and blame falls on them when AI goes wrong, organizations will either slam the brakes or push ahead in ways that generate backlash and quiet sabotage. The fact that nearly half of firms do not involve the CHRO in AI strategy is a red flag for anyone betting on smooth AI rollout.

For the race to AGI, this matters because systemic adoption is what gives advanced models real leverage in the economy. A world where workers distrust AI outputs and have no clear accountability framework will be slow to hand higher-stakes decisions to more capable agents. Conversely, firms that redesign workflows around explicit human-led, AI-assisted and AI-executed roles, as IBM highlights, are likely to capture outsize gains and build the internal norms needed to handle stronger systems later.

In practice, this report is ammunition for those arguing that AI strategy belongs at the board and C-suite level, with HR and operations as first-class stakeholders. Labs can ship increasingly capable models, but if enterprises fail to invest in critical thinking, override skills and shared accountability, they will choke off demand for frontier systems long before we hit AGI-level capabilities.