Frequently asked questions.

Altheris is AI agent security infrastructure. Wrap any AI agent in two lines of code and get 13 layers of real-time protection active immediately — no configuration files, no manual setup.

Install the Altheris SDK, then wrap your existing agent with it. All 13 security layers activate automatically — there's nothing to configure layer by layer.

No. Altheris wraps your existing agent stack — it doesn't require rearchitecting or rewriting your agent's logic.

Altheris has been tested across multiple AI models, including GPT-4, Claude, and Gemini.

Any AI agent — customer support, sales, financial, scheduling, data processing, and more. It's been tested across all of these agent types.

No. Altheris is built for teams without dedicated security budgets — it's designed to be handled by developers directly, not by a specialist security engineer.

Starter (£29/mo), Growth (£19/mo), and Business (£49/mo), plus an Enterprise tier for larger teams. Each tier increases the number of agents and features included — see the Pricing page for full details.

You can upgrade or downgrade your plan at any time from your account's billing settings.

Yes — you can cancel your subscription at any time from your billing settings.

Yes, Altheris offers a Free tier so you can try the SDK before committing to a paid plan.

Each plan has an agent limit built in — check the Pricing page for the exact number on each tier. Upgrading to a higher tier increases your limit.

Get in touch at help@altheris.io and we can discuss options for your specific situation.

Altheris addresses the OWASP LLM Top 10 and Agentic AI Top 10 — including prompt injection, excessive agency, credential abuse, memory poisoning, and more — through its 13 security layers.

Every action your agent takes is logged with decision provenance recording, giving you a full audit trail of what happened and why.

Altheris includes an emergency lockdown feature — the account owner can stop every agent instantly the moment a compromise is detected.

Altheris scans incoming instructions in real time, comparing them against known injection patterns and behavioural signals, and blocks anything that attempts to override your agent's original instructions.

Yes — rate limiting and cost circuit breakers are both active by default once the SDK is installed, protecting against runaway usage and cost spikes.

Email us at help@altheris.io and we'll get back to you as soon as possible.

Email help@altheris.io with details and we'll investigate promptly.

Yes — send suggestions to help@altheris.io. We read every message.