US enterprises prioritise AI but face major integration barriers

A new survey by Zapier indicates that most large enterprises in the United States are prioritising artificial intelligence, yet face substantial barriers to effective implementation.

The findings reveal that while 92% of enterprises treat AI as a priority, four out of five companies struggle to integrate AI tools with their existing legacy systems. This demonstrates a notable disparity between the aim to deploy AI and the ability to operationalise it within current organisational frameworks.

Integration challenges

The survey, conducted by Centiment for Zapier, gathered responses from more than 500 senior leaders at large companies. It found that 78% of enterprises have difficulty connecting AI solutions to existing systems, with a majority (53%) rating integration as at least moderately difficult. High vendor costs and concerns about security were also cited as significant challenges, with 45% pointing to high costs as a barrier, 38% questioning vendor security, and 33% expressing apprehension about potential vendor lock-in.

Organisational skill gaps were another major finding, with 35% of respondents reporting that a lack of employee expertise in AI technologies is hindering adoption. In terms of attitudes towards AI, 56% of enterprise leaders described themselves as "enthusiastic champions" of AI adoption, while another 28% are cautiously optimistic. A very small proportion (4%) are actively resistant, and only 2% said they had not yet considered AI adoption at all.

"There's a huge gap between wanting AI and actually making it work in complex enterprise environments," said Emily Mabie, AI Automation Engineer at Zapier. "Our survey shows the biggest barriers aren't about believing in AI's potential. They're about integrating AI with their existing systems, skill gaps, high costs, and vendor lock-in."

Competitive pressure

The findings also highlight the intense competitive environment surrounding AI. According to the survey, 81% of companies feel compelled to accelerate AI adoption to keep up with their rivals, while 41% reported that slow implementation has led them to fall behind competitors. Other reported consequences include missed productivity gains (39%) and delayed returns on investment (37%).

Infrastructure and data concerns

The study goes beyond fears around employee displacement, instead highlighting pragmatic barriers like data quality (29%), IT infrastructure limitations (27%), and difficulty quantifying returns on AI investments. These issues are hampering businesses' efforts to successfully deploy AI at scale.

There is also an observable divide regarding which departments lead AI initiatives. The IT function remains over 10 times more likely to spearhead AI acceleration compared to departments such as sales, marketing, human resources, or customer service. Yet, these leading departments can simultaneously become bottlenecks, with IT and C-suite teams often citing infrastructure constraints and approval processes as obstacles.

"The story here isn't about people resisting AI," Mabie added. "It's about organizations trying to manage a really complicated situation. They need to bring in transformative technology while handling security, governance, legacy systems, and teams that have different levels of AI knowledge."

The execution gap

The findings underline what Zapier calls an "execution gap." Investment in AI and pro-adoption attitudes are high, but many companies lack the infrastructure, integration capabilities, and cross-functional teamwork required to deploy AI at scale. This gap has led to setbacks in productivity, missed competitive opportunities, and delayed financial returns-a scenario recognised by a substantial proportion of enterprise leaders surveyed.

Zapier's recommendations to address these challenges include ensuring interoperability between AI tools within technology stacks, automating workflows across business functions, and integrating AI within unified company-wide processes. Additionally, the company advocates for democratizing access to AI tools across different teams, while maintaining adequate IT oversight for compliance and security.

The survey's respondents were exclusively U.S. C-suite executives, business owners, partners, and presidents at enterprises with over 1,000 employees. The results reinforce the view that while AI remains a top priority across large organisations, a range of technical, financial, and organisational challenges must be resolved to realise its full potential.