All spotlights

Product spotlight · March 2026

Set AI policy once for your whole organisation

AI governance lets you set organisation-wide AI policy in one place: model access, a guardrail preset aligned to Australian Government classifications, budgets and content filter policies.

New product AI governance For government and enterprise

What it includes

The March 2026 changelog has the full details.

  • Guardrail presets

    You can enforce a guardrail preset (OFFICIAL, OFFICIAL:Sensitive or PROTECTED) on every AI request. Each preset applies content safety filtering aligned to the Australian Government security classification of the same name. The preset overrides the guardrail settings of agents and requests.

  • AI access for the organisation

    You can turn AI services on or off for the whole organisation from one page. To open it, select AI Governance under AI Services in the dashboard sidebar.

  • Model policy

    You can allow or block specific models for the whole organisation.

  • Budgets and current spend

    You can set monthly and daily budgets for the organisation and for each user. The page shows current spend next to these settings.

  • Filter policies

    A Filter Policies section under AI governance lets you define content filtering rules for AI input and output. Make a policy mandatory for all agents in the organisation, or assign it to individual agents from the agent editor. A test panel lets you check a policy before you save it.

  • API and Go SDK

    Governance settings, spend and filter policies are also available through the v3 API. The Go SDK releases this month add the AI governance endpoints.

Who it is for

  • Government agencies

    Enforce an OFFICIAL, OFFICIAL:Sensitive or PROTECTED preset on every AI request your organisation makes.

  • Security and risk teams

    Choose the models your organisation can use, and make filter policies mandatory for every agent you run.

  • Budget owners

    Set monthly and daily AI budgets for the organisation and for each user, and see current spend on one page.

What it is

AI governance is a page in the Quant dashboard where you set AI policy for your whole organisation. It launched in March 2026.

Without it, each agent and each request carries its own guardrail settings. With AI governance, you set one policy, and it applies to every AI request in the organisation. On the same page, you control which models are available, set budgets and see current spend.

The guardrail presets use the names of the Australian Government security classifications: OFFICIAL, OFFICIAL:Sensitive and PROTECTED. Each preset applies content safety filtering aligned to the classification of the same name.

How it works

For example, an organisation that works with OFFICIAL:Sensitive information can set up the page like this:

  • Keep AI services on for the organisation.
  • Allow only the models that its security team has approved.
  • Enforce the OFFICIAL:Sensitive preset. An agent with its own guardrail setting still gets OFFICIAL:Sensitive filtering, because the preset overrides the guardrail settings of agents and requests.
  • Set a monthly and a daily budget for the organisation, and for each user.
  • Create a filter policy, check it in the test panel, and make it mandatory for all agents.

Filter policies apply to AI input and output. When only one agent needs a policy, assign it to that agent from the agent editor instead.

To turn AI services off for the whole organisation, use the same page.

Get started

  1. Open AI governance

    In the dashboard sidebar, select AI Governance under AI Services.

  2. Set access and models

    Turn AI services on or off, and allow or block specific models.

  3. Enforce a preset

    Select OFFICIAL, OFFICIAL:Sensitive or PROTECTED for every AI request.

  4. Set budgets

    Set monthly and daily budgets for the organisation and for each user.

  5. Add filter policies

    Test a policy in the panel, then save it. Read Filter Policies.

Related: Explore AI governance · Documentation · March 2026 changelog

Ready to try it on Quant?

Use AI governance on your own project, or talk to our team first.