Log in

A modern platform, protected by AI.

AI is changing how attacks happen. LeaseSide OS is defended with the same class of AI tooling, and it limits what its own AI is allowed to do.

Security at a glance

Guarded by AI agents

AI security agents review every change to the platform and check each release before it reaches production. An automated security audit runs every day.

Defended against prompt injection

Leases, emails and call transcripts are read as information, never as instructions, so a hidden prompt in a document cannot steer Clark. Microsoft's AI content safety also screens requests for jailbreak attempts.

Clark asks first

Creating, changing or sending anything needs a person's OK, and payments are off limits. The platform enforces this, not the AI's instructions.

Hosted in Australia

Your leases, documents and records are stored in Microsoft Azure's Australia East region.

Encrypted and firewalled

Encrypted in transit and at rest, with every request passing through a web application firewall.

Never used to train AI

Your documents and questions are not used to train any model, ours or Microsoft's.

How Clark reads your documents

Leases are information.

Never instructions.

Nothing sent without you.

Nothing used for training.

The questions a property team asks first

Where is our data hosted?

In Microsoft Azure's Australia East region. Your leases, documents and records are stored there.

Is it encrypted?

Yes. Every connection uses HTTPS, and Microsoft Azure encrypts stored data at rest with AES-256.

How is LeaseSide protected against AI-driven threats?

With AI of its own. AI security agents review every change to the platform and check each release before it reaches production, and an automated security audit runs daily. Inside Clark, text from documents, emails and calls is fenced off as information so a hidden instruction cannot take control, Microsoft's AI content safety screens for jailbreak attempts, and every AI request is recorded.

Who can see what?

Access is set by role. You decide who joins, what each role can see and who approves. Every change and approval is logged with who made it and when.

Can Clark act without us?

No. Clark prepares notices, tasks and replies, but nothing is created, sent or changed until a person approves it, and Clark cannot make payments. The platform enforces these limits, so a cleverly worded document or message cannot talk Clark past them.

Do you train models on our documents?

No. Your documents and questions are not used to train any model.

Which model providers are used?

Clark runs on OpenAI models provided through Microsoft Azure OpenAI, under Microsoft's enterprise data terms. Microsoft may process a request in another Azure region, but does not use it to train models. If the model is unavailable, the rest of LeaseSide OS keeps working and Clark shows as unavailable.

What certifications do you hold?

LeaseSide does not hold a security certification of its own yet, and we will say so until we do. The platform runs on Microsoft Azure, which is certified to ISO 27001 and SOC 2.

What is on the record

Every extracted field carries the document, page and clause it came from, the confidence at extraction, any correction made and the person who confirmed it. Every consequential action carries what was prepared, what it was based on, who approved it and when. Audit is not an add-on; it is how the product works.

Still have a question for our CTO?

Bring your security team to the first call. We will answer the hosting, access and AI questions in full.

Talk to Clark

Questions about security? Ask Clark.

Hi, I'm Clark. Ask me anything about where your data lives, who can see it and what I'm allowed to do.