No Jitter Brief:
Legacy systems weren’t built for AI agents
Two-thirds of organizations say legacy systems are the main reason they have problems scaling AI, according to the most recent Google Cloud/MIT report. Another 68% also said that legacy systems prevent agents from making fast decisions.
No Jitter
Publisher
Aug 25, 2026 at 5:37 PM UTC · Updated 2時間前 · 2 分で読める

Two-thirds of organizations say legacy systems are the main reason they have problems scaling AI, according to the most recent Google Cloud/MIT report. Another 68% also said that legacy systems prevent agents from making fast decisions.
About half of the respondents in the report say legacy systems have a negative impact on ROI. "Running AI agents with legacy data systems requires a lot of procurement, infrastructure cost, and engineers to maintain and develop them. That entails massive upfront investment," said Ashutosh Mishra, vice president, data strategy and platform, Deutsche Telekom.
"Legacy data was never built to be verified at scale, which AI requires. It was built to run one system, for one purpose, under rules that changed a dozen times and were never fully documented," Daniel Saroff, group vice president, research and consulting, IDC told No Jitter.
No Jitter Insight:
Related:Confluence POINT provides AI-enabled automation alongside existing tools
Legacy data is information stored in outdated formats, decommissioned computer systems, and obsolete or legacy software that is no longer supported by a vendor.
"Estimates based on surveys indicate organizations report 30% or more of their data is legacy data hosted in legacy systems," said Saroff.
Saroff said these systems weren’t built with AI in mind and that they have changed a lot over the years, sometimes without proper maintenance or documentation. He said there are 8 reasons legacy data disrupts AI models:
Article Intelligence
Topics
Sponsored
AdNewsLayer Premium
Unlock deeper intelligence.
Ad-free reading, exclusive research, and real-time onchain insights.
Go Premium
