AI is disrupting the talent landscaping on all sides.

While the tech has promised to boost productivity and automate tedious tasks, the rapid evolution of AI and its surrounding technology is making it difficult for enterprises to find the talent to keep up with it. The cost of this talent is also putting stress on IT hiring teams, said Brittany Lutes, research director of the CIO practice at Info-Tech Research Group. 

According to Info-Tech’s 2025 Talent Trends report, the top concerns for enterprise hiring managers are the cost of hiring adequate talent and the lack of capable workers. Even if a company does have the resources to spend on the best of the best, those employees are often few and far between, said Lutes. 

“If they’re not constantly staying up to date on the technical skills that they need … what they’re doing can become redundant,” said Lutes. 

The problem is that many people coming out of university right now – as well as those currently on the job market – are simply not equipped to meet the rapidly changing demands of enterprises, said Lutes. 

• The skill gap exists in AI and machine learning itself, as well as its peripheral technology, like cybersecurity, data analysis, compliance, cloud computing and infrastructure, she said. 
• “It takes four years to get a degree, but in that time, the curriculum that they would have built and what they would have learned would need to have changed twice by the time they graduated,” Lutes said. 

Because of this, companies are often turning to early-career programmers whose salary demands are lower and pairing them with AI coding assistants, while using senior-level developers to fix their mistakes, said Larry Kodali, CEO and founder of AI hiring firm OptimHire. 

While this allows enterprises to cut costs, the consequence is that mid-career developers are left in the lurch, he said. “Mid-level developers are really finding it really hard to find jobs,” said Kodali. “Repeatable, easy, low-end tasks – all of that could be done by fresh developers with these (coding) tools.”

This means that these employees need to upskill – and quickly, said Kodali. While there’s always been a scarcity of highly-skilled and senior-level developers, “now there’s much greater demand,” he said, especially as AI tools come into play. 

Mid-level job seekers either need to “upskill themselves faster for critical or hard-to-solve problems … or accept the new reality and start learning the new tools that are out there, and maybe join at the lower salary,” Kodali said. 

However, the onus of skill development may not solely be on these workers themselves, said Lutes. Enterprises can offer practical training and career development opportunities for early and mid-career employees to take advantage of, and “create environments where people can try things and explore and learn on the job.” 

Not only will this allow businesses to train for the exact skill they need, but it may be less costly than simply acquiring talent, she said. Retention tends to be cheaper than hiring, so companies will “probably spend a lot less money if you develop that individual and give them a career path than you would if you went to market,” she said. 

“Going forward, we need to stop trying to fill a (needs) gap with a role or a job title,” Lutes added. “What is it that we actually want to achieve, and what are we striving to go forward with in terms of organizational structure? I do believe development will be a better solution than talent acquisition.”

Tech is expensive. Enterprises are trying to figure out how to keep up. 

A Gartner survey released in January forecasts worldwide IT budgets to grow nearly 10% this year, topping $5.6 trillion, up from $5.1 trillion in 2024. The report projects data center systems and software will see the biggest increases, at 23% and 14% respectively. Though the cost of AI itself is coming down, a lot of excess spending is the result of companies navigating their cloud and infrastructure needs, said Bruce Kornfield, chief marketing and product officer at StorMagic.

Managing budgets is difficult when there are “a hundred different ways to design your IT infrastructure,” said Kornfield. But figuring out what capacity your business needs for cloud versus on-premise data storage is all about balance. “There are good use cases for cloud to run applications and store data, but when it comes to on-site, there is a need for it.” 

The question enterprises need to answer is where exactly those needs are, he said. While a lot of processing can happen in the cloud, finding a balance that satisfies both budget and processing needs is key, said Kornfield. “You can’t wait for analytics in the cloud,” he said. 

“The economy moves fast. The world moves fast, and real-time decision making and data processing and applications many times need to be run on site,” said Kornfield. 

Investing in and building strong on-premise infrastructure creates another costly expense, said Kornfield, especially as AI demand and adoption continue to trend upward.  

• According to Gartner, spending on AI-optimized servers is projected to double that of traditional servers in 2025 at more than $200 billion dollars. 
• “There’s a business need for many of these enterprises to start to build their own AI infrastructure on site,” he said. “(Businesses) need more processing power. They need GPUs … that is also driving the need for more budget.” 

But in a market flooded with AI products, it can be hard to decipher what’s actually worth the cost. The question to ask isn’t how to fit AI into your enterprise, but where it’ll actually make a difference, he said. 

“Throwing (money at) big, expensive, high-performance servers and GPUs and then saying, ‘let me go find my AI partner,’ that’s not going to work, and that’s going to be a waste of money,” said Kornfield. “Start with the business problem … then see what technology flows from it.”

Blockchain’s biggest strengths are transparency and immutability. Intel may want to put this tech to work. 

The company filed a patent application for “blockchains for security IoT devices” that essentially use blockchain as a means of recording and regulation who has access to devices in an Internet-of-Things network. 

The filing describes a system that would prevent unauthorized access by leveraging blockchain to provide a tamper-proof record of interactions, policies and device identities. Intel noted that this patent would be particularly useful in the context of devices for “remote sensing and actuation functions,” such as those in manufacturing or industrial environments.

Intel’s patent highlights a key use case for blockchain outside of cryptocurrency or digital assets: security. The immutable nature of blockchain as a recordkeeper makes it particularly useful as a “complementary tool for security, especially with IoT,” said Jordan Gutt, Web 3.0 lead at The Glimpse Group. 

Within security, blockchain technology can provide an “audit trail,” allowing IT teams and enterprises to track exactly who accessed what, and when, Gutt said. While Intel’s patent lays this out for IoT devices, this can have use cases in other areas of security, such as identity management, he noted. And as enterprises battle credential attacks in areas such as cloud, this use case could help them more easily track down breaches. 

However, while blockchain offers an “added layer of trust, it’s not a standalone fix for security,” said Gutt. Within security protocols, blockchain needs to work alongside other strong protections, he said. “Blockchain isn’t a silver bullet for anything. It just adds a decentralized trust layer that would complement the more traditional security measures.”