Terms of Service

By using this website or any free Arsenal tool, you agree to these terms. If you don't agree, don't use the site or the tools.

• Browse, read, and share links to any page on this site.
• Quote and reference our published research with proper attribution and a link to the source.
• Contact us through the provided channels.

• Scrape the site programmatically beyond reasonable browsing rates, or use crawlers or bots to harvest content at scale.
• Mirror or republish the site or its content as your own work.
• Use automated tools to harvest contact details for unsolicited outreach.
• Interfere with the site's operation, security, or infrastructure — including denial-of-service attempts, injection attacks, or unauthorized access to any backend systems.
• Use the site to transmit spam, malware, phishing material, or any unlawful content.

Free Arsenal tools are provided to help AI security teams improve their own systems. Open-source tools are MIT licensed — see each tool's repository for the specific license and what it permits.

• System prompts, documents, configurations, and test payloads that belong to you.
• Materials from systems you have explicit written authorization to test.

We process tool inputs through audited infrastructure with zero retention — inputs are discarded immediately after processing. See our Privacy Policy for the full details.

Free tools apply rate limits (typically a small number of scans per IP per hour). Exceeding rate limits results in a temporary block — not account termination. Rate limits exist to keep the tools available for everyone, not to push you toward a paid tier.

We may modify, discontinue, or limit any free tool at any time. For breaking changes we'll provide reasonable advance notice where possible. Tools are experimental — they reflect our current understanding of the threat landscape, not a static specification.

An account is required to save scan results, access usage history, and use the Pro tier tools. You can use the free Arsenal tools without an account.

• Keep your credentials confidential. Don't share your account with others.
• Notify us immediately if you suspect unauthorized access to your account: security@logicleak.io.
• You are responsible for all activity under your account.
• Don't create accounts through automated means or create multiple accounts to circumvent rate limits.

We may suspend or terminate accounts that violate these terms, with or without notice depending on severity. You may close your account at any time. On closure, your data is deleted within 30 days. Any scan results you have not exported will not be recoverable after closure.

• Site content — text, design, code, and branding — is owned by LogicLeak unless otherwise noted.
• Open-source Arsenal tools are MIT licensed. Each tool's repository specifies the exact terms. MIT is permissive: use, modify, and redistribute freely with attribution.
• Research publications are CC BY 4.0: you can reuse and adapt them with attribution and a link to the original.
• Brand assets — the LogicLeak name, logo, and product names — are not open-source. Using them commercially or in a way that implies endorsement requires our written permission.

You retain ownership of any inputs you submit to our tools or any messages you send us. We don't claim rights to your content. We process it only to run the tool or respond to your inquiry — see our Privacy Policy for details.

Our tools exist to help defenders. The line is authorization: testing a system you own or have written permission to test is fine. Testing anything else is not.

• Use any free Arsenal tool against systems you own or operate.
• Use any free Arsenal tool against systems where you have received explicit written authorization from the system owner to conduct security testing.
• Run our tools in lab environments, sandboxes, and non-production systems you control.
• Share tool outputs in security reports, research, and internal presentations.

• Test or probe any system — AI or otherwise — without explicit written authorization from its owner.
• Use our tools to bypass security controls on systems you don't own.
• Use our tools to facilitate unauthorized access to AI systems, model weights, training data, or proprietary system prompts.
• Run automated red-team campaigns at scale against production systems without documented authorization.
• Modify and redistribute our tools in a way that strips the authorization requirement from documentation or user-facing copy.

Unauthorized testing of computer systems may violate the US Computer Fraud and Abuse Act (CFAA), the EU NIS2 Directive, the UK Computer Misuse Act, and equivalent statutes in other jurisdictions. Any unauthorized testing you conduct using our tools is your sole responsibility. We cooperate fully with law enforcement when presented with valid legal process, and report observed misuse where required by law.

If you observe our tools being misused, or if you suspect someone is using them to target systems without authorization, contact abuse@logicleak.io.

Free tools are provided “as-is” and “as available” without warranties of any kind, express or implied. This includes disclaimers of:

Free tools are shared because the field gets safer when defenders have better instruments. They reflect our current detection logic, not a contractual security guarantee. If you need audited, warranted security testing with defined scope, deliverables, and professional accountability, that's what our paid engagements are for.

This warranty disclaimer applies only to the free tools and website. Paid engagements have their own warranties defined in each Statement of Work and governed by our Engagement Terms.

To the maximum extent permitted by applicable law, LogicLeak is not liable for any indirect, incidental, special, consequential, or punitive damages arising from your use of this website or our free tools. This includes loss of data, loss of profits, or business interruption — even if we were advised of the possibility of such damages.

Our total aggregate liability for any claim relating to the website or free tools — separate from paid engagements — shall not exceed £100. [DRAFT — confirm with counsel before publication]

• This limitation does not apply to liability arising from our own gross negligence or willful misconduct.
• This limitation does not apply to any liability that cannot be excluded or limited under applicable law — including mandatory consumer protection or statutory rights.

Liability for paid engagements is governed by the signed Statement of Work and our Engagement Terms, not by this section.

You agree to defend and hold harmless LogicLeak and its personnel against any claims, damages, and costs (including reasonable legal fees) arising from:

This indemnification obligation does not apply to claims caused by our own gross negligence or willful misconduct.

We may update these terms periodically. Substantive changes are logged in the change log at the bottom of this page.

For material changes — those that meaningfully affect how you can use the site or tools — we will provide reasonable advance notice. This may be a banner on the site, an email if we have your address, or both.

Continuing to use the site after a material change takes effect means you accept the updated terms. If you do not accept the change, stop using the site and tools before the effective date.

These terms are governed by the laws of England and Wales. Any disputes arising from these terms are subject to the exclusive jurisdiction of the courts of England and Wales, unless you have a non-waivable right under applicable consumer protection law to bring claims in your own jurisdiction. [DRAFT — confirm jurisdiction with counsel before publication]

We prefer to resolve disagreements directly before they become formal disputes. If you have a concern about these terms or how we have applied them, email legal@logicleak.io before pursuing legal action.

We track substantive changes below. Minor edits — typos, formatting — are not logged.