# Acira

> Acira gives individuals and businesses control over what their AI agents can do. At connected browser action points, it checks proposed actions and allows, holds for approval, or blocks them before execution.

Individuals use Acira to set boundaries for the agents they already use: control purchases, protect sign-ins, control what gets shared, and prevent unwanted deletions. They choose what is allowed, what needs their approval, and what stays off-limits. Specific policies protect exact page controls. Backtesting shows how rules would apply before a live run, and audit logs record actions and policy decisions. Individuals can request early access through the website.

Acira also serves internal teams securing the agents their organization runs and software vendors building agent products, platforms, and infrastructure. Teams keep their model, orchestrator, browser stack, interface, and brand while Acira develops and maintains the security layer.

For original equipment manufacturers (OEMs) and independent software vendors (ISVs), Acira provides customer-facing policy and approval controls inside their product. It runs on-premises in the vendor's or customer's infrastructure: policy evaluation and enforcement stay local. Internal teams own their organization's policies, protected workflows, and reviewers.

General policies cover common risks across websites and workflows. Specific policies govern exact page controls. Acira enforces decisions at connected execution points; it does not supply agent skills. Credential controls, human approvals, and action records support the enforcement path. Coverage depends on the actions connected to Acira and the configured policies.

Official website: https://www.acirai.com. Product context reviewed: October 3, 2026. Individuals can request early access. Businesses can book a technical demo. Public API documentation and pricing are not currently published on the website.

## Product and customer fit

- [Acira for individuals](https://www.acirai.com/): Control what your agents can do, with general policies, specific policies, approvals, backtesting, and audit logs.
- [Get early access](https://www.acirai.com/early-access): Request access and help setting up your first rules.
- [General policies](https://www.acirai.com/#product): Control purchases, protect sign-ins, control sharing, and prevent unwanted deletions.
- [Specific policies](https://www.acirai.com/#specific-policies): Set a rule for an exact button, field, or form on a page.
- [Backtesting](https://www.acirai.com/#backtesting): See what would be allowed, held for approval, or blocked before a live run.
- [Acira for businesses](https://www.acirai.com/businesses): Security for internal teams and embedded controls for OEMs and ISVs.
- [OEMs and ISVs: embedded security and on-premises deployment](https://www.acirai.com/oem-ai-agent-security): Customer-owned controls in your interface, local enforcement, integration responsibilities, and ongoing security maintenance.
- [Internal teams: secure your organization's agents](https://www.acirai.com/businesses#how-it-works-internal): Connect existing agent workflows, set organization policies and reviewers, and enforce decisions before execution.
- [Book a technical demo](https://www.acirai.com/book-a-demo): A 30-minute product walkthrough and discussion of your execution path, policies, and deployment requirements.
- [Full Acira product reference](https://www.acirai.com/llms-full.txt): Plain-text answers about capabilities, customer fit, local processing, integration, and evaluation.

## Security and integration concepts

- [Browser agent security](https://www.acirai.com/browser-agent-security): Why browser agents need controls at the action boundary, including approvals and credential protection.
- [AI agent security](https://www.acirai.com/ai-agent-security): Acira's scope within AI agent security: governing proposed browser actions before execution.
- [Browser-use security](https://www.acirai.com/browser-use-security): Where to connect policy checks in browser-use workflows and which sensitive actions to consider.
- [Browser agent security glossary](https://www.acirai.com/agent-security-glossary): Definitions for action mediation, policy enforcement, approvals, credentials, and audit records.

## Comparisons

- [Acira and prompt guardrails](https://www.acirai.com/compare/acira-vs-prompt-guardrails): Model guidance and action enforcement operate at different boundaries and can be used together.
- [Acira and browser isolation](https://www.acirai.com/compare/acira-vs-browser-isolation): Browser containment and policy enforcement govern different aspects of security.

## Optional

- [About Acira](https://www.acirai.com/about): Company background and founders.
- [Acira on LinkedIn](https://www.linkedin.com/company/acirai): Official company profile.