One in six report AI-related supply chain data exposure

By Elena Vasquez

One in six supply chain, logistics, distribution and wholesale professionals said their company experienced an AI-related supplier or customer data incident in the past year. The InstaLILY survey covered 250 managers and executives at companies using AI in daily operations.

InstaLILY, which develops AI agents for distribution, launched the survey through Pollfish on Oct. 9, 2026. The results reflect respondents’ reported knowledge and experiences, rather than independently verified exposures.

InstaLILY defined an incident as supplier or customer data being exposed, shared or accessed incorrectly through an AI tool. The report did not assess severity, so it cannot distinguish minor internal mistakes from more serious incidents.

How much sensitive data can AI tools access?

More than three-quarters of respondents, 78%, said their AI tools could access customer pricing and margins. Two-thirds, 67%, said the tools could access supplier pricing and rebate information.

Those access figures describe possible tool permissions, not independently confirmed misuse. They nevertheless show how much commercially sensitive information sits inside workflows where AI can generate quotes or act on contract and rebate data.

The survey covers supply chain, logistics, distribution and wholesale roles and does not isolate carriers or forwarders by transport mode. For distributors, wholesalers and logistics providers, it identifies two commercial exposures: a customer’s contract price and a supplier’s pricing or rebate information.

Exposure can damage relationships with trading partners and influence technology choices, according to the report. For shippers, carriers and forwarders using such systems, the commercial issue extends beyond security reporting.

How do partner questions affect purchasing?

About 63% said a supplier or customer had asked how their company protects its information. For 30%, those inquiries influenced a decision involving a software provider, vendor or AI tool. The survey therefore identifies data separation as a purchasing factor for businesses placing pricing, contract and rebate data into AI systems.

Adam Isenberg, head of distribution at InstaLILY, said: “Speed means very little if a faster answer costs you a relationship that took years to build.” He added: “AI working in distribution should know that one customer’s contract price never belongs on another customer’s quote. Buyers should make vendors prove that before they sign.”

Why does confidence coexist with overlap?

Despite the reported incidents, nearly nine in 10 respondents, 87%, said they were very or somewhat confident their AI tools kept supplier and customer information appropriately separated. Across the same sample, 63% acknowledged some overlap between supplier and customer data across their company’s platforms, including AI tools.

Where do current controls fall short?

Respondents most often blamed AI tools that lacked data-separation design, at 22%. Cost followed at 21%, while a lack of technical expertise drew 20%. The ranking puts product design ahead of budget in the reported control problem, while also showing that expertise remains a material constraint.

IBM’s separate 2025 Cost of a Data Breach report found that 13% of surveyed organizations had experienced a breach involving AI models or applications. Among those organizations, 97% reported lacking proper AI access controls. Breaches involving unauthorized “shadow AI” added an average of $670,000 to breach costs.

What will buyers require next?

Nine in 10 respondents said their employers would be more likely to select an AI tool specifically designed for distribution and supply chain operations if it offered stronger controls over supplier and customer data. Another 92% said data separation was a high or moderate priority for their company over the next year.

The report recommends testing AI systems against routine business scenarios, including whether a tool can prepare a customer quote without exposing confidential supplier pricing or contract terms. It also recommends establishing when human approval is necessary before AI systems act on information.

Over the next year, the 92% priority figure should give vendors able to document stricter separation more purchasing leverage. AI adoption is likely to remain conditional on demonstrable controls over customer and supplier data.