The landscape of artificial intelligence (AI) adoption is rapidly evolving, with a notable shift in the factors businesses consider to be the most significant barriers. As companies increasingly focus on Agentic AI—a branch of AI systems designed to operate autonomously and make informed decisions—the traditionally paramount issue of cost is now seen as secondary to the pressing concerns of data readiness, quality, and accuracy.
Recent research from Sinequa by ChapsVision underscores this transformation. In a survey of 100 enterprise decision-makers from both the UK and the US, only 10% cited cost as a barrier to adoption, marking a drastic drop in its importance. In stark contrast, 19% of respondents identified data accuracy and quality as their leading challenges, indicating a paradigm shift in the priorities surrounding AI investments. Nonetheless, despite this optimism—66% of organizations expect to see a return on their Agentic AI investments within five years—an unsettling 61% acknowledged that their data readiness remains inadequate.
For AI systems to deliver real value, particularly in an Agentic context, they necessitate timely access to reliable and contextual insights across the organization. This demand highlights a critical area of concern: nearly two-thirds of organizations admit to facing persistent data challenges that limit their capacity to harness AI effectively. Among those challenges, data security and compliance worries rank highest, impacting 67% of firms. Interdepartmental data silos (47%) and the sheer volume and pace at which data is generated (37%) further complicate the situation.
The increased demand for effective solutions like enterprise search systems, pivotal for managing knowledge retrieval, has become ever more pronounced. These “intelligent retrieval systems” have historically enabled enterprises to extract and surface knowledge from disparate data environments. Findings indicate that 66% of respondents view enhancing expertise in intelligent retrieval as essential for overcoming implementation hurdles associated with Agentic AI.
Jeff Evernham, Chief Product Officer at Sinequa by ChapsVision, commented on these findings, stating, “There is no doubt we’re racing toward an agent-driven future. As users grow more comfortable relying on AI for everyday tasks… the demand for similar capabilities in the workplace will only accelerate.” He emphasized that while the industry buzz may revolve around Generative AI, the fundamental importance of accurate and accessible data cannot be overlooked. He further noted that sustainable business transformations will hinge on continuous investment in data readiness and intelligent retrieval frameworks.
Comparative studies from other sectors reveal similar patterns, reinforcing the emphasis on data quality over cost. For instance, a survey conducted among global manufacturing leaders revealed that 44% remain cautious about Generative AI rollouts due to accuracy concerns linked to AI-generated misinformation. Despite recognising potential cost savings from AI, the industry’s leaders underscore quality as the cornerstone of successful deployments.
In recognising the ramifications of these data challenges, organisations reveal vulnerabilities that could thwart AI’s potential. According to a F5 survey, 72% of decision-makers reported poor data quality and scalability as significant barriers to AI deployment. The urgency to address these data governance issues reflects a wider trend identified by Deloitte, which revealed the complexities of managing continuous data movement and the integration challenges it poses.
Moreover, findings from Appen’s State of AI Report indicate a worrisome trend; data accuracy has declined by nearly 9% since 2021, with 86% of companies now needing to retrain or update their models quarterly to maintain performance. This persistent need for high-quality data only exacerbates the challenges businesses face in their AI journeys.
The overall picture is clear: businesses are recognising the pivotal role that data quality and governance play in successful AI adoption. As the demands of the AI ecosystem evolve, so too must the strategies employed by organizations to mitigate risks associated with data lag, complying with security protocols, and overcoming internal data silos. To truly capitalise on the potential of Agentic AI, sustained investment in robust data infrastructures and intelligent retrieval systems is not merely beneficial—it is essential.
Reference Map
- Paragraph 1, 2, 3, 4, 5, 6: Source [1]
- Paragraph 3, 4: Source [2]
- Paragraph 3, 4: Source [3]
- Paragraph 4: Source [4]
- Paragraph 5: Source [5]
- Paragraph 5, 6: Source [6]
- Paragraph 6: Source [7]
Source: Noah Wire Services
Subscribe to Industry Updates
