How businesses can close the AI engineering gap

AI skills shortage is a project killer

The lack of AI talent in business settings directly impedes AI deployments in several ways, including the following:

• Project abandonment. The "2025 AI Skills Report" by Pluralsight found that 65% of organizations abandoned AI projects due to lack of skills.
• Slow development. Businesses without the right AI skills in place are developing AI tools too slowly, according to McKinsey & Company research.
• Deployment and scaling challenges. Lack of skills also affect the ability of AI projects to be deployed and scaled. Gartner reported that less than half of AI projects make it to production, while McKinsey noted that only 1% of companies have scaled AI across their enterprises.

The cost of talent: hiring vs. contracting

Demand for AI talent is outpacing supply, creating a seller's market that increases the cost of acquiring qualified engineers. Businesses must evaluate if their AI needs warrant hiring, contracting or a mix of the two.

Hiring

Bringing a skilled AI engineer in-house requires a significant financial commitment beyond base salary alone.

• Compensation packages. Senior AI engineers earn an average base salary of $185,000 and, in some cases, considerably more.
• Hidden costs. Recruitment for these roles is uniquely expensive. The costs of sourcing fees, interview hours and the ramp-up period for a new hire could be substantial before the first line of production code is written.
• Employee benefits. Businesses might have to offer a wide range of financial and other benefits to a prospective hire.
• Retention challenges. With the intense demand and competition for AI skill sets, employee retention is difficult. A competitor might poach a new hire, offering better compensation.

Contracting

"Renting" talent using consultant firms or agencies can fulfill an immediate need, but comes with high hourly markups that can drain budgets if not closely managed.

• Premium rates. Highly sought-after consultants can command $300 to $500-plus an hour, while  junior consultants often bill at a rate of $100 to $150 an hour. For critical short-term interventions, "surge pricing" can reportedly push rates as high as $900 per hour.
• Agency retainers. For ongoing support, recruiting and contracting agencies can sometimes tack on monthly fees and retainers.
• Institutional knowledge. Businesses run the risk that contractors will leave the company and take the knowledge gained with them to another organization.

Methods to bridge the AI engineering gap

Though their execution remains inconsistent, businesses are pursuing several strategies to close the AI engineering gap, including the following:

• Full-time hiring. Businesses are allocating more money to find and hire engineering talent, as 63% of hiring leaders say it's more challenging to source candidates with adequate AI skills versus other tech roles, according to General Assembly's "State of Tech Talent 2025" survey. In addition, 75% of HR professionals say they're hiring too quickly without building viable pipelines.
• Skills-first hiring. HR leaders are increasingly taking a skills-first hiring approach that focuses on professional certifications rather than college degrees.
• Contract talent. IT organizations have tapped into contract and freelance talent for decades to supplement internal resources and are taking the same approach to AI skills.
• Internal upskilling. Businesses are upskilling and reskilling employees to better understand, utilize and build with AI.
• Educational partnerships. Businesses are recruiting directly from training facilities, including AI bootcamps.
• AI productivity tools. To boost overall productivity, businesses are using AI to learn AI.

4 steps to close the AI engineering skills gap

When formulating an effective strategy to address the AI engineering gap, consider the following steps:

1. Assess the entire workforce's AI skill level

Conduct a skills inventory across the business. As part of the assessment, determine which roles need to use AI tools versus which roles need to build AI systems.

2. Define the company's AI objectives and requirements

Identify what the business wants to build and accomplish with AI. Is the business, for example, deploying AI chatbots for customer service, building predictive models for inventory management or implementing AI-assisted coding across engineering teams? Each objective requires different skill sets. For example, a customer service AI deployment might need prompt engineering and integration skills. Or, an AI model typically requires data scientists and machine learning practitioners. Document the specific capabilities required and compare them against the company's current AI skills level to identify the talent gap.

3. Choose a strategy based on realistic constraints

Upskilling employees internally is the best approach to building a long-term competitive advantage in AI, but it's important to understand the limitations to securing the best AI talent possible, including the following:

• Time constraints. If the organization wants to achieve short-term results, the fastest wins can often come from immediate deployment of AI productivity tools. Using an AI-powered coding assistant, for example, can boost developers' productivity.
• Budgetary constraints. Outsourcing AI skills can potentially provide savings compared to hiring engineers. Contracts can specify delivery and financial terms that meet specific criteria, including budgets.
• Capability constraints. When specialized AI skills are needed without making a long-term commitment, contracted talent offers the highest AI proficiency rates. Research by Upwork found that 54% of freelancers surveyed report "advanced AI proficiency" compared to 38% of full-time employees. 

4. Combine approaches to acquiring AI skills

The most effective strategy is built on multiple approaches simultaneously -- for example, deploying AI tools now while simultaneously implementing a long-term upskilling program. It's critical to assess, iterate, and continuously evaluate which approaches are working and which aren't.

Sean Michael Kerner is an IT consultant, technology enthusiast and tinkerer. He has pulled Token Ring, configured NetWare and been known to compile his own Linux kernel. He consults with industry and media organizations on technology issues.