Free written AI risk assessment
Get a second set of eyes on your AI risk.
Send a short description of your AI system. I'll reply with a plain-English assessment of the risks I would check first: what could go wrong, why it matters, and what to do next.
No call required. No code access needed. No generic checklist dressed up as consulting.
Tell me what the AI does, who uses it, and what data it touches.
No system access, login, code review, or sensitive file upload required.
You get the risks I would look at first, not a vague "AI is risky" lecture.
You'll know whether deeper testing is worth considering.
When this is worth doing
If your AI touches customers, employees, or private data, start here.
This free assessment is best for teams using AI in places where a bad answer could create legal, financial, security, medical, compliance, or reputation risk.
Chatbots and support tools
AI that answers questions, handles complaints, summarizes policies, or responds directly to users.
Employee assistants
Tools connected to internal documents, policies, customer records, tickets, or company knowledge.
Agents and workflows
AI that can recommend, approve, deny, escalate, refund, route, summarize, or trigger next steps.
Request your assessment
Tell me enough to spot the likely weak points.
You do not need perfect answers. A few plain-language details are enough for a first read.
I'll reply by email, usually within 24 hours.
Want to see the shape of it?
Review an example assessment before you ask.
The report is meant to be direct and useful: the likely risks, why they matter, and what should be checked next.
Common questions
Simple answers before you send anything.
Do you need access to our AI system?
No. This free assessment is based on your description, your industry, and known AI failure patterns. If deeper testing makes sense later, access can be discussed then.
Is this a full security audit?
No. It is a first read. You'll get the biggest risks I would check first, not a complete red team assessment.
Do I need to be technical?
No. Plain English is better. Describe what the AI does, who uses it, and what could go wrong if it gives a bad answer.
What happens after the free assessment?
If the risk looks low, I'll say so. If deeper testing seems worthwhile, I'll explain what I'd test and why. You can decide whether to take the next step.
What kinds of risks do you look for?
Prompt injection, jailbreaks, data leakage, hallucinations, over-permission, broken guardrails, unauthorized commitments, and situations where the AI can be pushed outside its intended role.
Ready when you are
Send a few details. Get a clearer read on your AI risk.
The form above is enough to start. You do not need to prepare a formal brief, export logs, or book a call.