AI tools are gaining access to the pricing, rebates and account histories that underpin distribution relationships — and some businesses are already reporting problems controlling how that information is handled.

That’s according to a new report from workflow automation software provider InstaLILY AI. Titled “Shared Risk,” the report surveyed 250 manager-level and above professionals in supply chain, logistics, distribution and wholesale whose companies use AI in daily operations, including features embedded in existing software.

Among respondents, 78% said their employers’ AI tools can access customer pricing and margins, with the same share reporting access to customer purchase histories. Another 67% reported access to supplier pricing and rebate data.

Confidence in those tools’ safeguards was high: 87% said they were somewhat or very confident that AI keeps supplier and customer data appropriately separated. Meanwhile, 63% reported some or full overlap in the platforms housing those records. The report noted that shared platforms do not inherently compromise confidentiality when effective permissions are in place.

Still, 16% reported an incident during the previous 12 months involving supplier or customer information being exposed, shared or accessed incorrectly through an AI tool. The survey did not establish incident severity or whether information reached anyone outside the business.

The report describes a risk extending across business relationships: An inappropriate disclosure could reveal one customer’s negotiated price to another or expose confidential supplier terms.

Those concerns are influencing technology decisions. Nearly two-thirds (63%) said a supplier or customer had asked how its information was protected. Among that group, 47% said the inquiry affected a decision involving a vendor, software or AI tool.

Additionally, 90% said their employers would be more likely to choose AI built for distribution and supply chain operations over a general-purpose tool if it offered better control over supplier and customer data. Separating those records within AI systems was a high or moderate priority for 92% over the next year.

The Pollfish survey launched Oct. 9. Its findings reflect respondents’ knowledge and experiences, rather than independent audits or estimates for all businesses in these industries.

Get InstaLILY’s full report here.

MDM Analysis

NAW’s 2026 AI Adoption Index found that only 32.2% of respondents had a formal AI policy or governance framework. InstaLILY’s findings illustrate why that foundation matters as AI reaches commercially sensitive records. Distribution executives should require vendors to demonstrate how permissions hold throughout a quoting or account-service workflow, including what appears in the final output. Sales and purchasing leaders should help define those boundaries alongside IT. A useful test: Can the tool complete the task while withholding another customer’s contract price or a confidential supplier rebate? Industry-specific branding alone does not establish that capability.