Innovative Uses of AI in HR – A Cross-Functional Board Meeting

Meeting Details

In late September, 82 members of various i4cp Boards gathered virtually to discuss generative AI and HR.  This was designed to revisit and advance the discussion many of these same Board members had during the June 2023 all-Boards discussion on the same topic.  

During this September meeting, 94% answered that their organization had either already implemented or planned to implement (50% and 44% respectively) an internal instance of generative AI.  This figure is more than double what i4cp found just two months earlier in its global study of more than 1,500 organizations, and reflects the speed at which generative AI is catching on.

Please visit Is HR Already Behind in the AI Revolution? to view a copy of the i4cp report.  What follows is a written synopsis of the September all-boards meeting that was created using ChatGPT and then modified by i4cp to ensure accuracy and contextualization: 

AI vs. Generative AI - Katheryn Brekken, Senior Research Analyst at i4cp, clarified that generative AI is a type of deep learning that recognizes patterns within large data sets and predicts the next piece of data, while HR has been using machine learning, a type of AI, for many years. Kevin Martin, Chief Research Officer at i4cp, underscored the need for HR to educate themselves on generative AI, regardless of their organization's executive position on it.

Guardrails and Maturity: AI in HR - It is critical for organizations to have guardrails in place for employee use of generative AI in HR functions. Katheryn noted that only 23% of companies are currently experimenting with or using generative AI in HR, but this number is expected to increase significantly. 

Innovation and Adoption of Generative AI - Participants were surveyed about the level of generative AI adoption at their companies. Results revealed a relatively high adoption rate, especially in HR, and some suggested that more organizations would start adopting these models due to recent announcements.

AI Plan Implementation - Successful examples were shared, including IBM and IKEA using AI to automate manual tasks and retrain employees for new roles. Kevin Martin added that according to their research, about 50% of AI innovators are either implementing or preparing to implement such a plan, compared to only 1% of AI laggards. The discussion emphasized the potential of AI to reshape workforce planning and job roles.

AI in Recruitment: Overcoming Resistance and Ensuring Ethics – GE Appliances has experienced a 2X increase in candidate response rates to emails from the company after implementing generative AI in its recruitment process. Further discussion stressed the importance of training employees on appropriate behavior when using generative AI, and Katheryn highlighted the success of GE Appliances in providing this training. Cameron Hedrick of Citi shared the experience of confining all work related to generative AI to a secure sandbox for testing. Katheryn Brekken warned against using publicly available large language models due to potential frauds and misleading applications.

Chat GPT Adoption: Success Within Company - Marsh McLennan discussed the success of implementing an internal instance of ChatGPT within their company, which has seen high uptake from employees. The initiative was led by their CIO.  While initial communication was minimal, demand from employees was high. The company's approach was to position the tool as an enabler that would benefit all employees, with a focus on knowledge workers and consultants.

AI Implementation and Low-Code Discussion - The conversation continued with a discussion around the implementation and benefits of a new AI system. The Roberta model, an open-source, trainable language model, can read text and identify themes with 95% accuracy. The AI can also analyze large amounts of open-ended data, solving previously unsolvable problems. Roblox shared that the AI system has been implemented in their tier one and tier two support, with plans to expand to other tiers. This system has allowed them to delegate more routine tasks, freeing up employees to focus on more strategic work. In addition, the team discussed the use of low-code to no-code software development tools.

Skills Categorization and AI Integration Discussion - Kevin Martin inquired about prompt crafting training for the HR team members. Judy Albers, Director of Research Enablement at i4cp, noted that while there were many courses available, most were geared towards technicians rather than the average businessperson.

• Abhijit Bhaduri of Microsoft shared that they were categorizing skills into three buckets, including creating learning parts, curating content, and building skills. He also explained the potential of AI in HR and talent management, emphasizing the need for employees to understand AI applications and for the tech team to manage the core AI technology. Abhijit highlighted the opportunity in creating learning materials and suggesting career moves based on individual skills, which could be done through AI.
• Kevin Martin asked about the utilization of generative AI to advance Microsoft's ability to discover skills and capabilities within their workforce. Abhijit explained that Microsoft is developing a detailed talent architecture that AI can use for upskilling, reskilling, and career progression. However, he also highlighted the limitations of AI, particularly its inability to perform tasks that require human creativity and intuition.

Future Skill Needs and Work Helix - Kevin Martin shared an example from Cisco's people analytics team about their process for identifying and preparing for future skill needs, specifically in the case of UI/UX. They validated the need by analyzing internal and external job postings and found that project managers had a significant percentage of the necessary skills. They planned to develop these project managers further to meet the 80-100% skill requirement needed for a UI/UX developer. Kevin also shared that Workday Ventures introduced them to a company called Work Helix, which uses generative AI as its core technology.

AI Savings, Workforce Impacts, and Unknowns - The potential for significant financial savings through the use of AI technology was discussed, but also expressed caution about its impact on the workforce. Food Lion warned of possible repercussions from the workforce if these savings were achieved through layoffs or reductions in force. ServiceNow emphasized the need to ensure employees understand and embrace these changes to avoid resistance.

Large Tech Investing in Generative AI - Katheryn Brekken reviewed the recent investments made by large technology companies in generative AI. She highlighted that these companies are not only developing the technology for their customers but also for their own internal use, including employee training. She also mentioned that some companies are creating communities on slack to support users.

Reviving Motivation: AI for Process, Not Productivity

Highlighted was the lack of motivation among employees due to a lack of understanding of how certain initiatives can benefit them. Katheryn Brekken emphasized the importance of viewing generative AI as a systems enabler, rather than just a personal productivity tool. She suggested that the benefits of generative AI could be substantial when applied to end-to-end processes at a departmental or regional level. The aim should be to use generative AI to automate tedious tasks and drive processes, rather than building processes around the technology. She also touched upon the importance of considering life experiences, languages spoken, and other factors in skills inventories and competency assessments.

Embracing Change for Market Performance

Kevin Martin noted the potential for significant advancements through a systematic approach, particularly in creating a culture that embraces change along with the importance of leaders modeling desired behaviors to influence the culture positively. 

• This highlighted the need to prepare for the rapid evolution of AI and its integration with humans, as this could potentially cause stress and burden. The ultimate goal should be to create a change-ready culture that could enhance employee outcomes and market performance.