Scaling AI agents with trustworthy data

MIT Technology Review Insights published a sponsored report on August 12, 2026, based on a survey of 300 data and technology executives, examining how legacy data infrastructure is constraining enterprise AI agent deployments. The report was produced in partnership with Google Cloud, which hosts the full download.

The survey draws a sharp line between "data leaders" and "data laggards." Across all respondents, AI agents have access to an average of only 45% of company data, per MIT Technology Review. That figure drops to 30% or below at organizations classified as laggards, while a smaller cohort of leaders provides agents access to more than 70% of their data. The trust gap follows the same pattern: roughly half of all surveyed organizations say they trust the decisions their AI agents make, compared to 100% of data leaders.

The performance divergence extends to scale and speed. According to the report, 66% of data laggards say legacy systems limit AI agent scaling, and 68% say those systems prevent agents from making decisions quickly. Among data leaders, only 8% report either constraint — a gap the report attributes to those organizations having largely modernized their data estates ahead of agentic deployments.

The piece frames the challenge structurally: agentic AI requires real-time, frictionless access to data spanning supply chain, point-of-sale, and HR systems, in both structured and unstructured forms, with business context attached. Legacy infrastructure — even systems updated a few years ago — struggles to satisfy those requirements, the report argues. It also cites a Gartner projection that AI agents will augment or automate 50% of business decisions by 2027 as a timeline pressure point for organizations still running on older data stacks.

Looking ahead, the survey finds that 100% of respondents plan to be using agentic AI within two years, with 69% expecting to deploy it widely. The top-ranked initiative to enable that scale is improving access to both structured and unstructured data for agents. Enhancing data and AI governance with business context ranks second, while data leaders are also prioritizing automation of data management tasks.

One caveat worth noting: the report is labeled sponsored content, produced by MIT Technology Review Insights — the publication's custom content division — rather than its editorial staff. The disclaimer states it was not written by MIT Technology Review's editorial team and that Google Cloud is the sponsoring partner.