There are significant gaps in AI success for organizations, with 30% of those surveyed recognizing the impact on new revenue streams.
Orlando, Florida, April 30, 2026 /PRNewswire/ — Most organizations are moving beyond experimenting with artificial intelligence, but few realize its full value. A new study from Harvard Business Review Analytic Services sponsored by Appian found that while 59% of organizations (with some level of AI adoption) have deployed AI in production, the majority are currently focused on incremental benefits, prioritizing efficiency and productivity over top-line growth.
In particular, AI has the biggest impact on improving productivity, but it is not an enabler of growth. Respondents said that of the AI performance measures their organizations track, most impact productivity gains (64%) and operational efficiency (58%), while metrics like new revenue streams (30%) and ROI (35%) are among the least likely to improve. This represents a significant opportunity for organizations to use AI to achieve broader business outcomes and growth.
“Enterprises are at a tipping point. Instead of using AI to increase productivity, organizations must evolve to focus on growing the business. That’s where Appian comes in,” said Matt Calkins, CEO of Appian. “AI’s true potential can only be realized when it moves from a standalone tool to an embedded, revenue-generating worker. To get there, leaders must prioritize the foundational orchestration and rules-based guardrails needed to safely apply AI to high-impact work.”
AI still exists outside the flow of work
In most organizations, AI is not integrated into how business is performed and is used in parallel to operations, limiting its ability to drive higher-level business outcomes. While only 18% of respondents reported that AI is primarily integrated within workflows, the majority (34%) continue to use AI as a standalone tool alongside processes/workflows, a further 34% report combining both approaches, and 12% still report not using AI in processes/workflows at all.
Most companies think they can get some benefit from AI, but not yet at scale
While most respondents are seeing some return from their AI investments, only 16% report realizing a high degree of measurable value. The majority said the impact was moderate (33%), mild (36%), or not measurable (8%). Still, expectations remain high, as 86% agree that their organizations are seeking to realize more business value from the use of AI. While it is clear that AI is delivering some results, translating those results into meaningful and scalable business impact has proven difficult.
AI is embedded in workflows to deliver value
As organizations pursue AI strategies, their value is closely tied to how effectively AI is integrated into workflows and applied to operational operations. According to respondents, 71% of organizations that have incorporated AI into their processes are seeing good or moderate value from their efforts. At the same time, nearly three-quarters report significant benefits from modernizing legacy infrastructure/systems (76%), consolidating data sources (75%), and aligning processes/workflows across systems/applications (73%).
Legacy systems continue to limit the impact of AI
Nearly seven in 10 (69%) respondents agree that legacy systems are limiting their ability to scale AI across their enterprise. This reinforces the need for modernization and improved integration across systems and data. Siled or low-quality data (34%), lack of integration between systems (31%), and lack of AI talent/skills (30%) are also most commonly cited as barriers to incorporating AI into workflows.
Delay in introducing AI agents in core operations
The study also highlights differences in how AI agents are being applied across the enterprise. Organizations are more aggressively deploying AI agents in areas such as software development (35%), IT operations (31%), marketing and sales (26%), and customer service (25%). In contrast, core operational areas such as procurement (9%), manufacturing (10%), and supply chain (11%) have more limited agent adoption. In these areas, processes tend to be more complex and require greater control and consistency. Governance becomes important as organizations seek to extend AI into these environments.
Most organizations lack the guardrails needed to safely scale AI agents
Although 92% of respondents agree that rules-based guardrails are needed for AI agents to operate safely and effectively, fewer than half (48%) agree that their organizations have defined such rules (of those using, considering, or considering agent AI). As organizations consider agent AI systems (25% of organizations currently use them and 62% are considering them), the need for clearly defined processes and guardrails becomes even more important. Without clear guardrails, agents can behave unexpectedly throughout the system, increasing the risk of unintended consequences.
Process design emerges as key to unlocking the value of AI
To realize the full value of AI and achieve sustainable ROI, we need to rethink how work is structured and managed. According to respondents, organizations are increasingly focusing on better defining the rules/guardrails that AI must follow (50%), standardizing processes/workflows across departments (49%), and increasing collaboration between departments (47%) to increase the success rate of AI implementation.
”Although organizations are implementing AI, many have not integrated it into the core processes that drive business outcomes.said Alex Clemente, managing director of Harvard Business Review Analytic Services.Companies that successfully incorporate AI into their workflows will be in a position to realize meaningful value.. ”
Read the full study.
About research
In March 2026, Harvard Business Review Analytics Services, sponsored by Appian, conducted a survey of 385 business decision makers at organizations exploring, piloting, or actively using artificial intelligence (AI).
About Appia Island
Appian provides process automation technology. We automate complex processes for large enterprises and government agencies. Our platform is known for its unique reliability and scale. We’ve been automating processes for 25 years and understand business operations like no one else. For more information, please visit appian.com. [Nasdaq: APPN]
Logo – https://mma.prnewswire.com/media/1488235/5943388/Appian_Caption_2700px_Logo.jpg
