A1 Technologies launches Guardrails to bring unmanaged AI under security and governance

Australian Microsoft-focused managed service provider A1 Technologies has launched Guardrails by A1 Technologies, an AI security and governance service designed to help mid-market organisations across Australia and New Zealand manage risks from unmanaged AI applications and agents within their Microsoft ecosystems.

The service addresses risks that extend beyond formally deployed platforms such as Microsoft 365 Copilot.

A1 Technologies said employees across client organisations may already be using AI applications including ChatGPT, Claude and Gemini outside centrally managed environments, while organisations are also developing agents through platforms such as Copilot Studio and Microsoft Foundry - introducing risks around access permissions, sensitive data and oversight.

Guardrails runs as a quarterly cycle across four stages: discover (identifying AI applications and agents, reviewing overshared data, establishing AI system registers); govern (developing AI policies, acceptable-use requirements and agent lifecycle management); manage (using Microsoft Defender to identify and control AI applications, Purview to protect sensitive data, and Agent 365 to oversee agent identities and activity); and report (delivering executive reporting, risk reviews and agent lifecycle oversight).

Engagements begin with a fixed-price Survey that produces a Compliance Score - a measure of how much of an organisation's AI use is known, owned, controlled and recorded. A1 Technologies said it remeasures the score each quarter.

"AI use is moving faster than many organisations' ability to govern it," said Rob Rattray, director at A1 Technologies.

"Businesses need to understand what AI is already being used across their environment, what information it can access, and who is accountable for managing that risk. We've formalised this service because AI security and governance need to be managed as an ongoing business responsibility, not treated as a one-off technology project."

Clint Shiels, technical director at A1 Technologies, said the most common gap is governance and ongoing visibility.

"Organisations may have focused on the initial securing of an AI platform but have not defined a process to discover new or changes in AI use, who owns an AI platform for development and what that platform should be allowed to connect to," Shiels said.

"We see this as an ongoing compliance requirement to manage functionality and data security."

A1 Technologies said Guardrails draws on its existing Microsoft 365 environment hardening and Purview compliance work, extending that expertise to AI discovery, access permissions, data exposure and agent activity.