Skip to content
Halo by OmniTrust
New field guideBuild AI budgets that follow the owner
AI governance platform (AIGP)

Discover, secure, govern, and optimize AI where the work happens.

Halo finds the AI your people and agents use, applies policy in browsers, on laptops, at model and tool gateways, inside agent runtimes, and records each decision so reviewers can see what happened and why.

Book a walkthrough Explore the platform

Or start with a secure agent from $99 per month

Discover

See the AI your teams already use

  • AI sites, extensions, apps, and local runtimes
  • Sanctioned, monitored, shadow, and blocked
  • AI BOM export in CycloneDX, JSON, and CSV
Learn more about discover

Connects to the models, browsers, devices, and apps your teams already use

  • OpenAI
  • Anthropic
  • Google Gemini
  • Amazon Bedrock
  • Google Chrome
  • Microsoft Edge
  • Safari
  • macOS
  • Windows
  • Model Context Protocol (MCP)
  • Slack
  • GitHub
  • Jira
  • Notion
  • HubSpot
  • Zendesk
  • Snowflake
  • Datadog

One control plane

See it, secure it, prove it, and spend with intent

Halo brings discovery, runtime security, governance, and cost control into one continuous system.

Expose shadow AI

Find AI sites, extensions, apps, developer tools, and local runtimes on the browsers and devices you enroll, including the ones nobody approved.

Enforce at the moment of action

Policy decides before a supported tool call runs or a governed action goes through.

Hand reviewers real evidence

Decisions, approvals, and changes are recorded with reasons and export as CSV or JSON.

Spend with intent

Budgets act at thresholds you set, and cost shows up by agent and by model.

The product

A control plane your operators will actually use

Fleet health, quick deploy, budgets, and security decisions, all in the current Halo interface.

Halo Quick deploy screen. Set the job, model, apps, approvals, and budget together.
Set the job, model, apps, approvals, and budget togetherCurrent Halo interface

Coverage

Govern the AI you build, buy, and find

Whether an agent was purpose-built, arrived inside a tool your team adopted, or was never approved at all, Halo brings discovery, policy, and evidence to the supported surfaces where it runs.

Agents you build

Launch agents with security built in

Set the job, model, app access, approvals, and budget together, then run in a sandboxed runtime with brokered credentials.

Learn more about launch agents with security built in

AI your teams already use

Bring shadow AI into the inventory

Halo Browser and Halo Endpoint surface AI in use on supported surfaces and classify it, so security can decide with facts.

Learn more about bring shadow ai into the inventory

Models and tools your agents call

Govern each model request and tool call

Route model and MCP traffic through gateways that apply identity, policy, approvals, and budgets, and record the outcome.

Learn more about govern each model request and tool call

The Halo stack

Discover, secure, govern, and optimize, working as one

From the browser tab to the budget report, each layer shares identity, policy, and evidence with the others.

  1. Budgets and routingSpend limits that act on their own and route profiles for each workload
  2. Evidence and analyticsDecisions, approvals, and costs recorded and exportable
  3. Policy and identityPolicy as code, agent identities, and human approvals
  4. Gateways and runtimeModel, MCP and tool, and API gateways plus a sandboxed agent runtime
  5. Browsers and devicesHalo Browser and Halo Endpoint discover the AI in use

Why Halo

Built for how AI actually shows up at work

Halo connects visibility, enforcement, and evidence, so the rules you write are the rules that run across the supported browsers, endpoints, gateways, and agent runtimes your teams use.

Enforcement where AI actually runs

Browsers, devices, gateways, and agent runtimes each get controls suited to them, under one policy model and one evidence record.

Security and efficiency in one decision

The same request is checked for policy, budget, and model fit, so safer AI is also the more economical path.

Simple enough to start this week

Launch a secure agent through self-serve plans, or start discovery in observe mode with a guided pilot.

Browsers supported
3
Chrome and Edge, plus Safari on macOS
Desktop platforms
2
macOS and Windows
Gateways
3
Model, MCP and tool, and API
Catalog apps
1,000+
Through the Halo connection catalog

Halo connects to the tools your teams already run

Model providers, browsers, devices, MCP servers, and 1,000+ apps available through the Halo connection catalog, each one governed with scoped access and evidence.

See governed integrations

For your team

Built for the teams that own AI risk and value

Security, IT, governance, AI platform, and finance leaders each get the view and the controls they need.

Security leadersFor CISOs and security architectsStart with AI policy enforcementIT and platform teamsFor CIOs, IT, and platform engineeringStart with shadow AI discoveryGovernance, risk, and complianceFor AI governance and risk leadsStart with AI compliance evidenceAI buildersFor AI platform and application teamsStart with AI agent governanceFinance and AI operationsFor owners of AI spend and valueStart with AI cost governance

Ways to start

Start with one agent or govern the whole organization

Self-serve

Put a secure agent to work

Plans from $99 per month include the secure runtime, governed model and tool access, budgets, and audit evidence.

  • Halo-managed and BYOK model access
  • App connections from the catalog
  • Budgets and approvals included
Start with HaloSee pricing

Enterprise

Govern AI across the organization

Guided discovery on browsers and devices, policy at scale, a security review, and support terms that fit your requirements.

  • Guided Halo Browser and Halo Endpoint rollout
  • Custom policies and approvals
  • Custom DPA, security review, and audit support
Book a walkthrough

From the blog

Field guides for governing AI

All articles
AI Efficiency5 min read

Build AI budgets that follow the owner

AI budgets work when each agent has an owner, thresholds trigger clear actions, and exceptions leave a record. See how account and per-agent budgets fit together.

Read article
AI Governance5 min read

Who reviews the governance assistant?

Assistants that draft governance work need provenance, permission boundaries, and an accountable human decision. Learn how to review their suggestions before anything is approved.

Read article
Compliance & Evidence5 min read

What ISO/IEC 42001 asks of your AI software

ISO/IEC 42001 sets requirements for an organization's AI management system. Learn what auditors may ask of your AI tools and what evidence software can supply for that system.

Read article

FAQ

Frequently asked questions

Halo Browser for Chrome and Edge, plus Safari on macOS and Halo Endpoint for macOS and Windows identify AI websites, extensions, desktop apps, developer tools, local runtimes, and add-ins on enrolled devices, then organize them into an inventory with source and confidence.

Agents that run on Halo get a sandboxed runtime, brokered credentials, policy, approvals, and budgets. Model and tool traffic that flows through Halo gateways receives the same identity, policy, and evidence controls.

Policy evaluates the request at runtime. Halo can allow it, deny it, raise an alert, or pause it for a person to approve, and it records the decision with the reason so reviewers can see what happened.

Gateways move traffic and dashboards report after the fact. Halo combines discovery on browsers and devices, gateways for model and tool traffic, a secure agent runtime, policy, budgets, and evidence in one control plane.

Account and per-agent budgets notify, pause new operations, or stop active deployments at thresholds you choose. Cost analytics show spend by agent and model, and model routing profiles help match each workload to an approved model that fits its cost and quality needs.

You can sign up and launch a first secure agent through the self-serve quick deploy path. Browser and endpoint discovery rolls out with guided onboarding, usually starting in observe mode before teams turn on warnings or enforcement.

Put a halo around the AI your business runs.

See discovery, runtime policy, approvals, and evidence working together on your own use cases.

Book a walkthrough Start with Halo