Adoption challenges, accessibility, and AI bias

People technology investments can fail long before a system goes live — not because of technical faults, but because of how people interact with, access, and are affected by those systems. This video addresses three of the risks listed under AC 3.2 of FREE02: low adoption, accessibility barriers, and bias in data and AI. It examines why organisations routinely underestimate the human dimension of technology implementation and how each of these risks can quietly erode the value of even well-resourced deployments.

The video covers the specific causes behind low adoption, including inadequate training, poor usability, and absent leadership endorsement, before turning to the consequences of uneven system engagement on data completeness and reliability. Accessibility barriers are treated as a distinct risk category, with the video addressing how disability, digital literacy, language, and device access intersect with the design and rollout of people technologies — and situating this within equality legislation rather than treating inclusion as optional good practice. The discussion of AI bias focuses on how algorithmic systems trained on historical organisational data can perpetuate and amplify demographic disparities in decisions such as hiring, appearing objective whilst producing discriminatory outcomes.

Algorithmic bias mitigation, disparate impact testing, human oversight of automated decisions, and the conditions under which employees should be able to challenge AI-driven outcomes are all topics the video addresses in the context of a people professional's responsibilities. Taken together, the three risk areas covered — low adoption, accessibility exclusion, and AI bias — form a substantial part of the indicative content for AC 3.2, making this video directly relevant to anyone preparing to discuss the risks and limitations of people technology and analytics within the FREE02 unit.