Ai laptop hero

To Build or Buy AI skills?

Job postings including generative AI as a skill have increased while relatively few organizations are upskilling their workforce

Wielding generative AI (GenAI) competently is becoming more important to every organizational function—evidenced by nearly every tech platform in existence announcing new AI features.

Generative AI may seem simple enough to use informally. But organizations won’t benefit from these new tools, or worse, they could open themselves to lawsuits or cyber attacks, without some basic skills and education within its workforce to:

  • Effectively prompt models (which are AI systems trained to generate or predict information);
  • Employ digitally safe behaviors; and
  • Prevent these models from disadvantaging under-represented populations.

To prepare for integration of GenAI in the workplace, new data shows more organizations are buying, seeking external candidates with AI related skills instead of training their own.

This article explores:

  • Why organizations need more AI-related skills
  • The current state of organizations’ efforts to build and/or buy these skills
  • The costs and benefits to either strategy

Why organizations need more AI-related skills

Nearly a quarter of companies (24%) have an internal generative AI model for the enterprise (e.g. an internal large language model (LLM) trained on the organization’s data) according to data from a recent survey of 1,332 participants conducted by the Institute for Corporate Productivity (i4cp). Another 13% are planning to implement one.

GenAI, if used properly, gives companies huge advantage over competitors that don’t use the technology in terms of increased productivity and time and cost savings.  

A slightly smaller number (21%) of those surveyed reported that their organizations encourage employees to use publicly available models to achieve these benefits.

Enterprise wide use of generative ai

These organizations need a workforce equipped with certain AI-related skills. If companies aren’t providing employees who are eligible to use GenAI with training in safe digital practices, they’re putting their sensitive data at risk and opening the opportunity for phishing and other criminal cyber attacks. And if they’re not providing employees with the knowledge to properly prompt these models, they’re not going to realize the advantages of this technology to productivity and quality improvement.

The state of organizations’ efforts to build and buy AI-related skills

An increasing number of organizations, but still a relatively small proportion, are developing at least some of their employees to effectively leverage GenAI.

Developing Workers to effectively leverage generative ai

Doing so has positive benefits; i4cp’s report Is HR Already Behind the AI Revolution? found that upskilling the workforce to effectively and safely leverage generative AI has a cascade of benefits, including higher market performance, productivity, innovation, and healthier cultures.

This i4cp-identified next practice—one that a small percentage of early adopters are pursuing (and are likely to report having high-market performance)—is necessary to create a culture of safe digital behaviors to protect companies’ sensitive data.

Only a minority (22%) are doing this, and it’s a slightly smaller group than those that have implemented an enterprise GenAI model.

Furthermore, most organizations that are developing their workers to use GenAI are only upskilling a portion of their total workforce.

Who at your organization is offered training related to generative AI

Only 10% of those surveyed from large organizations (those with 1,000+ employees) reported that training related to GenAI is being offered to all employees. Survey participants at organizations that offer training to all were more likely to indicate that their HR is using GenAI and that the organization has high confidence in HR to lead on AI strategy (no doubt because HR’s L&D teams successfully orchestrated the training for everyone).

Instead of upskilling their workforce, companies are increasingly seeking these skills in the job market, as indicated by new data from Revelio Labs, a workforce analytics company.

In the last six months, the number of U.S. job postings mentioning ‘generative AI’ increased 53% over the number of requisites mentioning this skill the year prior

In the last six months, the number of U.S. job postings mentioning ‘generative AI’ increased 53% over the number of requisites mentioning this skill the year prior. Revelio Labs looked at 10,374,508 job postings between February and September and compared the skills sought to those in 25,351,480 postings between September 2022 and 2023.

Between Sept. 2022 and Sept. 2023, the total job postings mentioning generative AI was 5,073, but since then there have been 7,787 postings with that skill listed.

Top Trending AI-Related Skills in Job Descriptions

AI Skill

% Increase in total job postings referencing skill

Generative AI






Large language models


% Increase represents the change in total number of U.S. job postings mentioning that skill between Sept. 2023 and Feb 2024, compared to the total number of U.S. job postings mentioning that skill between Sept. 2022 and Sept. 2023. Skills definitions: Generative AI uses deep learning to recognize patterns in data to predict the next word, pixel, etc.; TensorBoard is a suite of visualization tools, created by Google, for machine learning experimentation; OpenAI is an AI research company that provides services such as ChatGPT; and large language models predict and generate human language. Source: Revelio Labs

Interestingly, a separate analysis of just HR job postings from Revelio Labs found the number of HR job postings referencing AI-related skills has nearly tripled since 2019.

The costs and benefits to either strategy

Considering GenAI’s potential impact to payroll alone, organizations shouldn’t rely on just hiring to build a workforce with the right skills to leverage these new tools.

Implementing a new AI strategy, especially one involving an enterprise-wide generative AI model, necessitates careful consideration of the costs and benefits associated with either approach.

Upskilling the existing workforce


  1. Retention of institutional knowledge: Current employees already understand the company's culture, processes, and goals. Upskilling them ensures this knowledge is retained and applied in the context of new AI tools.
  2. Employee engagement and satisfaction: Investing in employees’ development can lead to increased engagement and productivity.
  3. Cost-effectiveness in the long term:  While there are upfront costs associated with training, upskilling may be more cost-effective over time compared to the expenses of recruiting, hiring, and onboarding new employees.
  4. Ensure ethical use and inclusivity: Current employees might be more invested in the company's ethical guidelines and the importance of using AI in a way that does not disadvantage protected populations. Tailored training can reinforce these values.


  1. Time and resources for training: Developing and delivering a comprehensive training program requires significant time and resources.
  2. Temporary decrease in productivity: As employees undergo training, they may have less time in the short run to carry out their responsibilities.
  3. Skills gaps: Depending on the existing skill levels, there might be considerable gaps that require extensive training. Some employees might struggle more than others, necessitating differentiated learning paths which can complicate training efforts.

Hiring Externally


  1. Immediate access to expertise: Hiring externally can bring in individuals with the exact skill set needed for working with generative AI models. This can speed up the implementation of AI strategies.
  2. Bringing in fresh perspectives: New hires can bring innovative ideas and perspectives that might not be present in the existing workforce, potentially leading to more creative uses of AI within the company.
  3. Addressing skill gaps quickly: External hiring allows the company to specifically target and fill skill gaps, ensuring that all necessary competencies are covered.


  1. Recruitment and onboarding costs: Finding, hiring, and onboarding new talent can be expensive and time-consuming.
  2. Potential for salary compression: Specialists in AI and related fields might command higher salaries. This may impact salary schedules, particularly for roles in which incumbent employees doing similar tasks are just now developing AI-related skills or don’t have them at all.
  3. Blowback from stakeholders: When companies replace workers due to shifting business priorities (recall IBM’s 2016 ‘rebalancing’ strategy), they can face significant criticism from current employees, customers, or other stakeholders. This can diminish trust between employees and leadership and even harm revenue streams.   

Ultimately, the decision about upskilling, hiring externally, or a combination of these, depends on several factors, including the current skill level of the workforce, the urgency of the AI strategy implementation, the available budget for training versus hiring, and the company's long-term goals.

For most cases, a hybrid approach will be most effective, combining the benefits of retaining and developing current talent with the infusion of new skills and perspectives from external hires. This approach allows for the most balanced and strategic implementation of AI capabilities within the company.

Katheryn Brekken, Ph.D.
Katheryn Brekken, Ph.D., is a senior research analyst with the Institute for Corporate Productivity (i4cp). Prior to joining i4cp, she served as an assistant professor of research with the MGM Resorts Public Policy Institute at the University of Nevada, Las Vegas School of Public Policy and Leadership, where she continues to lecture. She has worked closely with government and corporate leaders to develop and evaluate education and training programs and as a policy advisor. She has over 15 years of experience in public affairs and has testified before legislative bodies on matters of higher education and workforce policy. She is published in various academic journals including Politics & Policy, Community College Journal of Research & Practice, and State and Local Government Review. She received her Ph.D. in Public Affairs from UNLV.