Organizations increasingly expect agentic AI to fundamentally change how work gets done, promising significant gains in productivity and growth. A new report suggests that few companies possess the necessary processes or workflows to realize these benefits. While the technology is advancing rapidly, the internal structures required to support it remain largely underdeveloped.
According to the study, about half of surveyed leaders believe they understand how AI agents will impact their future operating models. Yet, adoption is constrained by several persistent barriers. The primary challenges include a lack of a unified data foundation, limited trust in AI governance, and the high cost and complexity of integrating these systems into existing infrastructure.
Fewer than half of these respondents feel their organizations are prepared for agentic AI across most business areas. Workforce readiness and business processes rank as the weakest points. It suggests that while businesses may acquire the tools, they still need to develop new ways of working alongside them.
Related: Wireshark 4.6.8 fixes 28 security flaws, nine in parsers
Laura Shact, U.S. Technology, Media and Telecommunications (TMT) AI growth leader at the firm, noted that superficial changes will not suffice. “AI transformation includes moving beyond one-off fixes to a sustained focus on improving work outcomes,” Shact said. “Limited, layered-on approaches may create the sense of getting ahead with quick wins, but in reality, they may not be enough.”
She added that thriving in this future will require more than just alignment with technology strategies. It demands investment in work and organization design, as well as intentional leadership and workforce enablement.
Processes Remain the Primary Bottleneck
The data reveals a stark gap between ambition and readiness regarding operational workflows. Only 16% of leaders say their organizations’ processes are fully ready for agentic AI, with a mere 5% describing them as highly prepared. Even among organizations that have already deployed AI agents at scale, readiness is limited.
Related: Anthropic asks for testing instead of bans
Only 46% of these advanced companies report that their business processes are adequately prepared. Executives attribute this lag to several practical issues. Many cite poorly documented processes, fragmented data systems, and entrenched working habits. There is also a noted shortage of AI expertise among both executives and employees.
Just one in five organizations believe they are prepared to redesign their business processes to operate autonomously with AI agents. Current conditions resemble the enterprise resource planning implementations of the 1990s, where companies often tried to automate existing inefficient processes rather than optimizing them first, leading to expensive failures.
History suggests that simply layering advanced technology over outdated workflows often amplifies inefficiencies rather than resolving them. The current hesitation to rebuild processes from the ground up follows a similar pattern of prioritizing short-term functionality over long-term structural integrity.
Related: New cybersecurity tools hit the market
Many organizations are currently introducing AI agents into existing workflows rather than redesigning processes from the ground up. This “layered” approach offers a quicker path to short-term returns and helps build experience. However, long-term value is expected to depend on a complete overhaul of business processes around the technology.
The transition is expected to take time.
