Feature
While businesses race to declare their commitment to AI, many are putting their money into areas where the technology doesn’t help business.
Just 1% of business leaders see their AI deployment efforts as “mature,” but 92% intend to increase their investment in AI over the next three years, according to McKinsey. A recent MIT study found that 95% of generative AI pilot projects fail to deliver any kind of measurable returns. The problem isn’t the technology, but a lack of integration and learning, as well as the difficulty of incorporating tools like ChatGPT into corporate workflows.
As for the 5% of projects that do in fact succeed, they have a common element: focus. The companies involved “pick one pain point, execute well and partner smartly with companies who use their tools,” MIT researcher and lead author Aditya Challapally told Fortune.
Match AI Skills to the Work
AI works best with back-office automation, MIT’s study found, meaning it excels at tackling repetitive and administrative tasks. But when it comes to actually applying AI, about half of the funding is invested in pilot projects that use the technology for sales and marketing chores — activities where human participation is important and the focus is on people rather than machines.
Similarly, a separate McKinsey study found 42% of companies use AI for marketing and sales. Twenty-eight percent use it for work in product and service development, while about 23% use AI in IT. Just 13% use it for HR.
Some companies make things worse by attempting to develop AI solutions in-house, even though the talent for running such projects remains relatively rare. Only a third of AI tools developed in-house (33%) met expectations, compared with two-thirds of those built by specialized AI providers. This is particularly an issue with financial, healthcare and other regulated companies, where that failure rate puts their compliance at risk.
Add one more factor: Human workers are still cheaper than AI. Citing an earlier study from MIT, SiliconAngle reported that only 23% of wages could be cost-effectively replaced by AI in roles that require some kind of computer vision skills, such as teachers, appraisers and bakers.
Even when they do work, corporate efforts in AI generate only modest returns. While 70% reported gains in revenue because of using AI in strategy and finance, nearly half of the companies surveyed (47%) reported increases of 5% or less, McKinsey said. Just 11% saw increases of more than 10%.
Focus on the Need
Organizations seeing the best results “are the ones that are thinking in terms of wholesale transformative change that stands to alter their business models, cost structures and revenue streams — rather than proceeding incrementally,” observed Alex Singla, senior partner of McKinsey’s AI consulting practice.
They do this by sweating the details. AI’s greatest impact comes from lower-level efforts such as redesigning workflows rather than higher-level functions, like writing advertising copy.
Even when other systems and workflows are properly aligned with AI’s capabilities, companies set themselves up for problems. As a quote widely attributed to Bill Gates goes, “The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.”
So why do executives still try to use AI in areas where it might not make much of a difference? Some feel pressured to keep up with competitors while others simply don’t have the technological knowledge to understand what they’re doing. Others hope AI will deliver quick business wins or short-term advantages in the market. For many, it’s simple FOMO — the fear of missing out.
They may also be thinking at too high a level. Saying “AI isn’t meant for use in marketing and sales” doesn’t mean AI is unsuitable for every job involved in marketing and sales. For example, Accenture used cloud-based AI to improve delivery of marketing analytics. Its SynOps platform compiled data into one view, streamlined content production and eliminated 55% of the manual steps needed to manage a marketing campaign.
“The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.”
Where AI Makes Sense in HR
So how do we make AI work for HR? Start with a manageable, high-impact problem where automation delivers quick wins, such as reducing the time spent on repetitive work like drafting policy documents or summarizing data.
Doing this accomplishes two things. First, success early on builds your team’s confidence and demonstrates real value, especially when you select projects with clear KPIs that track outcomes and impact. This helps build your case for future AI budgets.
It’s also important to involve colleagues and team members in your work. Bring middle managers on-board by sharing the purpose and long-term impact of your AI efforts. Training and coaching improve the team’s confidence, and teach managers and staffers to balance new digital tools with human abilities such as emotional intelligence and ethical practices.
Conference
Sep
16
HR Tech Conference Las Vegas 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
Finally, don’t go it alone. Working with IT and AI professionals helps make sure your efforts integrate with existing systems.
By piloting projects first, scaling after your theories have been proven and gaining some early wins, HR leaders position AI as a supportive tool that augments rather than replaces people.
Related Articles:
- How HR Leaders Bridge Human Potential and AI Innovation — Four ways HR leaders can rise up to the challenge and integrate technology and the employee experience.
- What Happens When Candidates Choose AI Over Recruiters? — A fact that should make every talent leader sit up: when over 70,000 applicants were offered a choice between a recruiter and an AI interviewer, 78% chose AI.
- Better Work or Better Cover? The Double Life of AI Use in the Workplace — We're using AI to prop up a culture that prizes being quickly right over slowly wiser. Here's what to do to make sure it accelerates human potential.
