Amid pressure to deploy generative and agentic solutions, a familiar question is surfacing: “Is there an AI bubble, and is it about to burst?”
For many organisations, this new wave of generative and agentic AI is still very much in experimental stages. The primary focus, and the low-hanging fruit, has been internal. Most businesses are looking to AI to increase efficiency gains, such as automating workflows or streamlining customer support. The trouble is, those gains are proving elusive.
Ben Gilbert, VP of 15gifts, points out that “those benefits often take years to show real returns and are hard to measure beyond time savings.”
This is where the cracks begin to show. The rush to deploy feels uncomfortably familiar and, for some, may give some feelings of PTSD.
“The trend of companies diving headfirst into AI projects or solutions mirrors patterns we have seen time and time again in previous tech bubbles, such as the dot-com era,” explains Gilbert.
This gap between experimental spending and measurable profit is precisely where the bubble is weakest.
Gilbert argues that AI projects which “focus on efficiency gains and deliver unclear or delayed ROI” will be the first to fail from any bubble pop. When investments “risk becoming costly experiments rather than profitable tools,” the pullback is inevitable.
“We could see budgets tighten, startups close, and large enterprises re-evaluate their AI strategies,” says Gilbert.
It’s a warning backed by data. Gartner has already predicted “that over 40% of agentic AI projects will fail by 2027 due to rising costs, governance challenges, and lack of ROI”.
So, what separates a viable AI strategy that could survive a burst bubble from a costly experiment? Gilbert suggests it comes down to human nuance; something many projects overlook in the rush to automate. There’s a curious discrepancy, he notes: “Why has AI been embraced so fully in efficiency gains and customer support, but not in sales?”.
The answer may be that algorithms are highly valuable for sifting through data to inform decision-making, but consumers want the engagement, intuitiveness, and fluidity of human interaction as well. Success, then, isn’t about replacing people but augmenting them.
Gilbert advocates that “AI should be taught by real people, so it can understand the nuances of human language, needs, and emotions”. This requires a transparent process, where “human annotation of AI-driven conversations can help to set clear benchmarks and refine a platform’s performance.”
A total AI bubble pop isn’t likely to be imminent. Gilbert explains we’re more likely to see a “market correction rather than a complete collapse” and the underlying potential of AI remains strong. However, the hype will deflate.
For enterprise leaders, the path forward requires a return to first principles. “AI projects, whether built on hype or business value, need to address a real human need in order to be successful,” Gilbert says.
Whether a bubble or healthy market correction, this cooling-off period might even be a good thing, offering a chance for businesses to focus on AI quality over hype and smarter ethics. For the CIOs and CFOs managing the budgets, Gilbert believes the brands that thrive “will be the ones using AI to enhance human capability; not automate it away.”
“Without empathy, transparency, and human insight, even the smartest AI is destined to fail.”
See also: Keep CALM: New model design could fix high enterprise AI costs
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security Expo, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
