🔒CT experts see cautious, uneven progress as companies confront AI’s growing pains

One thing that generative artificial intelligence has no difficulty producing is splashy headlines. The latest have focused on major layoffs, as tech companies prune their workforce, in part, to reflect AI-driven productivity gains. By some counts, the tech sector has announced as many as 1 million job cuts in 2025, including recent reductions of 14,000 at Amazon and 24,000 at Intel. Yet, these developments contrast sharply with surveys by leading universities and research firms showing widespread caution — and even confusion — among business leaders about how to actually implement AI. Experts describe this as a field that’s changing month to month and quarter to quarter, generating a lot of speculation about how profoundly AI may reshape business organization and employment.

Learning gap

Stamford-based Gartner, a business and technology insights company, produces a series of reports on technology adoption called Hype Cycles. Its 2025 report on procurement and sourcing solutions places generative AI squarely in what it calls the “trough of disillusionment” — the phase when early excitement gives way to disappointment as companies struggle with underperformance, poor integration and unmet expectations. Another Gartner survey found that 88% of human resources professionals said their organizations have yet to realize significant business value from AI tools. Only 8% of those respondents believe their managers have the skills to use the technology effectively. Meanwhile, a widely cited MIT study found that 95% of corporate generative AI pilot projects have failed. “The 95% failure rate for enterprise AI solutions represents the clearest manifestation of the GenAI Divide,” the report stated, pointing to a “learning gap” affecting both the makers of AI tools and the organizations trying to adopt them. That learning gap is evident in Connecticut as well. A recent survey by the Connecticut Business & Industry Association found that only 36% of companies have begun using artificial intelligence in their operations, while 64% have yet to adopt it. Among non-users, 73% said they were unsure how to integrate the technology, 20% cited data privacy concerns and 7% pointed to liability risks. Among Connecticut firms that are using AI, applications include research, data analysis, project management, customer service, marketing, sales, human resources, accounting and communications. Some companies said they are using AI tools to help write emails, improve operational safety, support building design or enhance internal processes through platforms such as Microsoft Copilot.

Data security

The figures on partial adoption sound familiar to Wei Chen, a professor at UConn’s School of Business who has recently published a paper on generative Al and organizational structure in the knowledge economy. “Companies are still at the stage of choosing their tools,” he said. And beyond the significant investment in AI tools and the steep employee and organizational learning curve, there are larger structural issues for many businesses. “You just hear the same question again and again,” he said. “‘I have a data privacy problem. I’m a finance company, I’m a healthcare company, my data cannot go out of my company. Can I still use this technology?’” The data security question is holding many companies up from committing to fully integrating generative AI. But Chen says at the individual level, AI has proved a roaring success. “ChatGPT’s annual revenue is about $15 billion, and 70% of that comes from individual users who pay $20, $25 a month,” he said. “The individual adoption is here, everyone is using it. The organizational adoption is just starting.” Chen said the Gartner and MIT findings represent just the first, early wave of AI business adoption, and another, more successful wave will be coming as artificial intelligence firms better tailor their products to business needs — developing custom tools that protect company data and address integration and management challenges. He believes full adoption in many industries may take five to 10 years.

Running ahead

Business process consultants are among those on the front lines of advising companies on this technology revolution.

“In some ways, the impact of this change has similarities with earlier technology disruptions and breakthroughs,” said Dan Spiwack, CEO of Stamford-based business consultancy JMW, which advises corporate leaders on strategy, culture change and performance improvement, with clients that include Fortune 500 companies and government agencies. “(But) in my view, AI may potentially have an even greater impact in the world of business.” Getting there won’t be smooth. Spiwack said he frequently sees companies establishing AI teams or departments with sizable budgets but no clear strategy for how to deploy the technology — or how it will ultimately benefit their customers. In part, these moves are fueled by very human concerns — leaders either hoping to leapfrog the competition or afraid of falling behind if their competition gets to those applications first. “In some ways the remarkable capabilities of AI seem to currently be running ahead of companies’ abilities to apply those capabilities for differentiated performance,” Spiwack said. He advises that there’s no substitute for thoughtful, human-directed strategy. “We believe that it’s the human element of leadership that determines if those technology tools will be applied to actually produce exceptional performance,” he said.

Career ladder

Ultimately, for managers and employees alike, one of the biggest questions surrounding AI is its potential impact on organizational structure and headcount. In other words, will the widespread layoffs now occurring in the tech sector soon spread to other industries? Not everyone believes the tech layoffs actually represent solely the rise of AI, but instead reflect a mix of factors, including a potentially softening economy and a headcount correction after a pandemic-driven hiring wave in the industry. In fact, Gartner has commented on the Amazon layoffs saying they likely reflect a strategic “talent remix” to focus on high-priority business areas rather than directly resulting from AI-driven productivity gains. Chen’s research focuses on the continuing need for human intervention in the AI loop to reduce risks from hallucinations — the tendency of generative systems to produce information that sounds plausible but is factually incorrect or entirely fabricated. His models suggest productivity improvements from GenAI yield distinctive organizational shifts: as productivity increases, firms tend to employ fewer but more knowledgeable workers. But he acknowledges that raises fundamental questions for both businesses and society. If Chen’s predictions hold true, many industries could see the traditional career ladder start to disappear — with fewer entry-level roles and a greater premium on specialized, highly skilled positions. “If they don’t hire entry workers anymore, as a whole society, how does that work?” he asks. “I don’t think anyone has a real framework for how we should deal with this yet.”