Dive Brief:
- An AI success gap is emerging across enterprises, driven by a failure to fully integrate the tech into workflows, according to a Harvard Business Review Analytic Services survey commissioned by Appian published last week. The report surveyed 385 businesses using AI.
- To see true value-add, the report said businesses need to embed AI into operations from the ground up, and implement guardrails to safely guide the tech in high-level workflows. Among organizations embedding AI into processes, 71% reported moderate or substantial value, compared with an average of just 16% among the entire sample.
- "The true potential of AI can only be realized when it moves from a standalone tool to an embedded worker that drives revenue,” Matt Calkins, CEO of Appian, said in a release accompanying the report.
Dive Insight:
AI has firmly moved from an experimental tool to a core tenet of operations, yet few businesses are using it to its full potential.
A key issue, according to the report, is how the technology is being deployed. More than half of the organizations surveyed are already using AI in production, though most deployments are focused on efficiency gains rather than revenue growth.
“Enterprises are at an inflection point,” Calkins said. “Organizations must evolve to focus on business growth."
The findings come as hyperscale cloud providers pour hundreds of billions into AI infrastructure, rapidly expanding data center capacity and specialized chips to meet anticipated demand. While the spending boom reflects ongoing confidence in AI’s long-term economic impact, day-to-day returns remain uneven, and a majority of businesses are running AI alongside existing processes rather than integrating it efficiently within them.
At the same time, risk concerns are shaping deployment strategies. Nearly all respondents said AI agents require rules-based guardrails to operate safely, but fewer than half (48%) have established them.
Lack of AI governance has already emerged as a key concern for companies scaling the technology, with major vendors — including Salesforce, AWS and Databricks — releasing in-house tools designed to bridge this gap.
Legacy infrastructure is another stubborn barrier. Nearly 7 in 10 respondents said outdated systems are limiting their ability to scale AI, with data silos, integration gaps and skills shortages also highlighted as major obstacles.
For CIOs to effectively leverage tools for high-value use cases, third-party analysts including Gartner advise businesses to prioritize targeted use cases tied directly to measurable business outcomes.
KPMG also highlighted weak governance as a key factor causing AI pilots to fail. Without data readiness and controls, enterprises can’t effectively move the tech beyond experimentation.
As Appian's report found, moving toward higher-value use cases will likely require three efforts advancing at once: modernizing core systems, integrating fragmented data and redesigning workflows with AI embedded from the outset.
