Data Infrastructure, Not AI Models, Limits Agent Success
A MIT Technology Review Insights report based on a survey of 300 data and technology executives finds that legacy data systems are a major blocker to AI agent adoption and effectiveness. Organizations with mature data infrastructure, termed 'data leaders,' report significantly higher trust in agent decisions and fewer scaling constraints than 'data laggards.' The research suggests that without modernizing data systems, enterprises will struggle to realize ROI from agentic AI despite widespread adoption plans.
TL;DR
- AI agents currently have access to only 45% of company data on average, dropping to 30% or less in lagging organizations but exceeding 70% in data leaders
- Only half of surveyed organizations trust their AI agents' decisions, while 100% of data leaders report full trust, indicating data quality directly correlates with agent reliability
- Two-thirds of data laggards cite legacy systems as limiting agent scaling and decision speed, compared to just 8% of data leaders
- 100% of respondents plan to deploy agentic AI within two years, with 69% expecting widespread use, creating urgent pressure to modernize data infrastructure
Why It Matters
As organizations race to deploy AI agents, a critical gap is emerging between technical capability and operational readiness. The survey reveals that data infrastructure, not AI models, is the limiting factor for most enterprises. Without addressing legacy system constraints, companies risk deploying agents that lack the data access and context needed to make reliable decisions at scale.
Business Impact
For executives evaluating AI agent investments, the research shows that ROI depends primarily on data modernization, not agent sophistication. Organizations that delay data system upgrades will struggle with agent accuracy, speed, and scaling, potentially wasting significant capital on agent deployments. Data leaders are already capturing competitive advantage through faster, more trustworthy agent operations.
Key Implications
- Data infrastructure modernization should precede or accompany AI agent deployment, not follow it, to avoid costly rework and underperformance
- Organizations need to prioritize unified access to both structured and unstructured data across enterprise systems, including supply chain, point-of-sale, and HR data
- Data governance frameworks that embed business context are becoming operational requirements, not optional compliance measures, as agents make autonomous decisions
What to Watch
Monitor how quickly enterprises move to modernize legacy data systems in response to agent deployment timelines. Watch for emerging data infrastructure vendors targeting the agent-readiness gap, and track whether organizations that delay data modernization report lower agent adoption success rates or higher failure rates in production deployments.
Subscribe to the newsletter
The latest stories and analysis, delivered to your inbox.
Free. No spam. Unsubscribe any time.
