AI in production surges, but data quality lags behind : Apryse

ByRicardo

New international survey reveals 64.5% of enterprises have AI in production, but solely 38.1% fee their doc data as “glorious,” exposing a important hole in AI readiness.

Apryse, the document-to-data basis powering clever automation, at present launched findings from its international survey on AI adoption and doc infrastructure. The outcomes reveal a putting paradox: whereas AI has gone mainstream, most enterprises are nonetheless battling the doc data governance wanted to scale it.

Traditional data shared between paperwork tends to be messy, inconsistent, and exhausting for AI to interpret. While people can manually vet the data, that is unsustainable for many firms. Without clever pre-processing, these paperwork stay unstructured, making automation and correct insights more and more difficult. 

The September 2025 survey of 465 organizations throughout North America, Europe, Australia and New Zealand uncovered key insights: 

  • 64.5% of organizations already run AI in production; largely for enhancing operational effectivity (63%), enhancing the client expertise (51%), and making data-driven choices (41%). 
  • 76.6% retailer between 25–75% of their data in paperwork, but solely 38.1% fee that data as “glorious” for AI use.
  • 67.3% say preserving doc processing in-house is “extraordinarily vital.” 54% cite data safety issues as the highest barrier to scaling AI, and 49% cite data quality. 
  • 82.8% plan to speculate in doc automation throughout the subsequent 12 months—but practically half lack confidence in their present pipelines.
  • 62.8% expertise doc quality points “often” or “ceaselessly.” 

“AI is not experimental, it’s operational,” mentioned Andrew Varley, CPO, Apryse. “But enterprises are discovering that the infrastructure behind it, particularly round doc data quality, hasn’t advanced quick sufficient. Surging data progress with out governance, an absence of visibility into what content material already exists, and fragmented tooling are actually the most important boundaries to clever processing at scale.”

Asia-Pacific Quietly Leads in AI Maturity

While North America leads in AI deployment (77.7%), Australia and New Zealand quietly outpaced the West in infrastructure maturity. These respondents reported the best adoption of generative and predictive AI, hybrid cloud utilization, and OCR applied sciences, signaling a worldwide shift in innovation.

“Oceania is outpacing the West in a number of key areas of AI infrastructure,” added Varley. “The area has been an early-adopter of data residency guidelines and regulatory mandates, pushing organizations to embrace hybrid cloud and superior doc processing. With highly-regulated industries like healthcare, authorities, and monetary providers dominating the market in Oceania, this urgency has created a stable mannequin for correct document-to-data workflows.”

From Chaos to Context: The Need for Structured Document Data

The survey additionally highlights a rising demand for instruments that transcend digitization. Organizations want options that extract which means and construction from paperwork, not simply textual content. When requested about probably the most important capabilities in doc automation, respondents ranked:

  • Table/kind recognition (59.6%) to resolve the issue of understanding format and relationships in paperwork like invoices, contracts, and varieties.
  • Developer-friendly SDKs to decrease technical boundaries to automated doc workflows. 
  • Metadata tagging to allow context-aware data classification for compliance, searchability, and governance, lowering danger and enhancing AI accuracy.

Apryse’s embeddable SDKs and clever pre-processing applied sciences are constructed to satisfy this demand. By remodeling unstructured paperwork into structured, AI-ready data, Apryse permits scalable and clever processing with out compromising data sovereignty. To learn the complete report, go to apryse.com.

The submit AI in production surges, but data quality lags behind : Apryse first appeared on AI-Tech Park.

Similar Posts