Don’t Rely Solely on AI

Sixty-five percent of managers use artificial intelligence at work, reveals a 2025 Resume Builder survey of more than 1,340 US managers. Among them, 94 percent rely on it to assess direct reports, with the majority using it for performance-related decisions such as raises, promotions, layoffs, and firings.

The phenomenon reflects a broader corporate race to harness AI's promise and capabilities. A Business Insider article, "Amazon Now Requires Staff to Show AI Use to Get Promoted at Ring and Some Other Units," details how certain companies require employees to demonstrate AI proficiency for promotion consideration, making tools such as ChatGPT, Copilot, and Gemini essential performance benchmarks.

Additionally, one out of five managers lets AI tools make final decisions without human oversight. It's a slippery slope for organizations, Resume Builder reports, because, while AI can support data-driven insights that enhance productivity and scalability, the technology can't incorporate context, empathy, and judgment. Without robust training frameworks and clear guidelines, companies risk distorting talent pipelines, undermining trust, and harming organizational culture, according to the study.

For instance, a 2024 University of Washington study found that AI-based resume-screening tools favored White-associated names 85 percent of the time. Additionally, AI systems only selected woman-associated names 11 percent of the time, and the tools never chose Black-male-associated names over White-male-associated names.

Resume Builder also notes that AI models are only as fair as the data they process. Without proper insight into bias mitigation, employee recommendations can reinforce systemic inequities.

For the talent development industry, the business world's reliance on AI signals an urgent need to recalibrate learning and leadership programs. Resume Builder's survey found that two-thirds of managers using AI have not received formal AI training, only one-third received structured ethical coaching, and one-quarter didn't have any training at all.

KPMG's 2025 Trust, Attitudes, and Use of Artificial Intelligence study found that, while three out of five workers now use AI, only about half had any form of training, and many individuals don't evaluate outputs for accuracy.

As an example of a training initiative with positive results, The Guardian's "From AI Ethics to Inclusive Leadership: How Professional Development Gives Businesses an Edge" highlights the University of Melbourne's professional development courses. The program aims to address gaps in AI ethics and inclusive leadership, noting that modern leaders often overlook AI's bias traps without proper training. The solution, according to the article, is for TD practitioners to embed AI ethics and bias literacy into leadership programs, promote human-in-the-loop governance, and build transparency and auditability into training.