Expert shares four-point plan for employers to maximize AI
As organizations navigate the complexities of integrating generative AI (Gen AI) tools like ChatGPT into their HR operations, many leaders are searching for actionable strategies.
Brandon Roberts, ServiceNow’s GVP of people analytics and AI, recently outlined a structured four-point plan designed to help HR teams maximize AI’s potential while addressing challenges such as governance, adoption, and capacity building.
Currently, many employers have identified the use case on AI. However, once employers get there, things like governance, operating model, and other different components provide an obstacle, he says in talking with HRD.
“I think a lot of people are stuck there right now and trying to figure out how you continue to make progress on this.”
While artificial intelligence (AI) promises a boost in productivity for employers, it has yet to fully deliver, according to one expert. And that’s because employers are struggling to implement an enterprise-wide adoption of the technology, according to another expert.
Establish a robust operating model: A clear operating model is the foundation for successful AI integration. Roberts emphasized the importance of a streamlined process to evaluate and implement AI use cases across the organization.
“You need a way to say, ‘Hey, I have an idea or a use case that an organization is interested in… I need a way to intake those, evaluate them quickly, and get them through the right checks and balances,’” he explained.
Roberts’ organization has developed a centralized approach where employees can submit use case ideas, which are then routed to relevant technology teams. These teams assess the feasibility and value of each idea, conducting legal, ethical, and data security reviews before presenting them to a steering committee. Out of 550 submitted AI use cases, 14 were implemented—a hit rate that underscores the importance of rigorous evaluation.
Focus on enablement: Roberts highlighted enablement as a critical driver for adoption. Organizations often underestimate the need for comprehensive training to ensure employees can use AI tools effectively.
“Every time you launch a use case, you need to train employees on how to use it, or it won’t get adoption,” Roberts said.
He shared an example of an AI-powered case summarization tool launched earlier this year. Despite its ability to save agents 37% of their time, adoption was initially low due to a lack of training.
“We implemented training after we launched and said, ‘Hey, this is how it works, this is how you should use it.’ Adoption more than tripled,” he noted.
Roberts’ organization conducts quarterly AI training sessions for all 26,000 employees, ensuring foundational knowledge of AI tools and their appropriate applications.
Prioritize targeted upskilling: To fully realize AI’s benefits, organizations must upskill employees whose roles are impacted by AI-driven efficiencies. Roberts explained that when AI creates capacity in certain roles, it’s essential to redirect that time toward high-value tasks.
“How do we upskill those employees... to use that 10% capacity to do something meaningful with it that our organization cares about?” he posed.
This targeted approach to upskilling ensures that AI-driven productivity gains translate into tangible organizational value, rather than simply reducing workloads.
Build a strong data and technology foundation: High-quality data is the cornerstone of effective AI deployment. Roberts emphasized that organizations with poor data quality or weak governance frameworks struggle to derive meaningful value from AI.
“If people haven’t had good data historically, they’re getting no value from AI,” he warned.
To address this, Roberts’ organization has invested heavily in data management, ensuring that AI models are built on reliable and accurate inputs. This includes creating governance frameworks to oversee data privacy, security, and ethical considerations.
Recently, the federal government has launched the Canadian Artificial Intelligence Safety Institute (CAISI) to bolster Canada’s capacity to address AI safety risks,
Roberts underscored the need for a strategic and centralized approach to AI, cautioning against fragmented solutions that lack long-term scalability. He pointed to common use cases such as generating job descriptions and interview questions, which save significant time when powered by external AI tools.
However, enterprise-wide applications require more robust enablement, upskilling, and governance to succeed.
By implementing an operating model, focusing on enablement, investing in upskilling, and strengthening data foundations, Roberts believes HR leaders can turn AI’s potential into tangible enterprise value.
Policies surrounding the use of AI appear to be falling behind the growing use of the emerging technology among HR professionals, according to a previous report.
