A recent report from autonomous‑AI platform provider Teradata reveals that the biggest obstacle to deploying truly autonomous artificial intelligence across major U.S. industries is not technology but data. Healthcare, manufacturing, retail and financial services all report that their data is too fragmented, overly sensitive or lacking the contextual metadata that AI agents need to act reliably.
Healthcare lags behind on trustworthy data
Because health records are stored in a patchwork of electronic health‑record (EHR) systems, 90% of healthcare leaders say only 20% or less of their data is sufficiently described for AI agents to use. The average across all surveyed organizations is 77%.
Privacy and compliance concerns lead 42% of health‑care respondents to deliberately limit AI agents, requiring a human to execute the final action. As a result, 83% of health‑care firms remain in the experimenting or developing stages, with AI limited to low‑risk tasks such as billing. Only 2% have reached the operationalizing stage, the lowest rate among the four sectors.
Manufacturing sees strong interest but faces legacy‑system gaps
Manufacturing executives are optimistic: 87% view autonomous AI as a competitive advantage for supply‑chain optimization and predictive equipment maintenance. Yet only 8% have moved to the operationalizing stage.
The chief barrier is connecting modern cloud‑based services with older manufacturing control systems. Fifty‑four percent of leaders cite data‑connection issues as a top impediment, and a shortage of AI and data specialists further slows progress.
Retail struggles with siloed consumer data
Retailers hold vast amounts of consumer information, but that data lives in separate e‑commerce platforms, point‑of‑sale systems and third‑party services. Forty‑eight percent of retail leaders measure AI ROI primarily through customer‑satisfaction metrics, while 46% focus on engagement and service.
Because fragmented data hampers functions like autonomous inventory management and personalized shopping experiences, only 5% of retailers have reached the operationalizing stage, the second‑lowest rate among the surveyed industries.
Financial services balance risk and reward
Banking and financial institutions place the highest emphasis on enterprise‑wide ROI (65% of respondents). Yet the sector’s complex regulatory environment makes fully autonomous decision‑making especially risky. A mis‑step could trigger a failed audit or a costly loan error.
Half of financial‑services leaders say governance, security or access restrictions limit AI agents’ ability to reach the data they need. Despite a relatively even spread across the experimenting, developing and building stages, only 7% have achieved operationalizing status.
Common challenge: data and context fragmentation
The study, which surveyed 1,000 senior technology and data leaders from companies with at least 500 employees in the United States, United Kingdom, France, Germany, Japan and Saudi Arabia, concludes that a lack of unified, contextualized data is the single biggest roadblock to autonomous AI. Across all sectors, a quarter of organizations are in the building stage, and just 7% have reached the operationalizing stage where AI can deliver enterprise‑wide ROI.
Teradata’s “Arrested Automation: Why Agentic AI Stalls at the Enterprise Level” highlights that before companies can reap the promised returns, they must first invest in data governance, lineage and enrichment to give AI agents the reliable context they need.
Original reporting: KEYT (Ventura/Santa Barbara) — read the source article.
