Your product runs on AI. Your contracts should say how.
For AI and SaaS founders in Toronto and across Ontario. We write the terms your customers sign, the privacy policy that matches what your models do with data, and the agreements that keep the code in your company. Every piece of work has a fixed fee, agreed in writing before we start.
What we help with
- AI product termsTerms that say who owns what your users put in and what your model puts out, and whether you can train on it. From $760
- SaaS or master services agreementA customer contract that holds up when an enterprise buyer sends it through procurement. From $920
- Terms of serviceRules for your app or platform that match how it actually works, not how a template assumes. From $500
- Privacy policyA privacy policy that tells users plainly what your product and its models do with their information. From $500
- Founder vesting and IP assignmentThe code, prompts and designs you made before incorporating, assigned to the company, with vesting so a departing founder doesn’t walk off with equity. From $920
- Consulting and contractor agreementFor the consultants, freelancers and dev shops building with you, with a written IP assignment so the code comes back to the company. From $450
- NDASomething to sign before a design partner, vendor or prospective hire sees your model, data or roadmap. From $300
- SAFE roundYour pre-seed or seed round on SAFEs, including the securities filing. From $2,600
- Trademark search and filingYour company or product name, protected in Canada before you scale the brand. From $920
Prices shown are starting prices in Canadian dollars for the base scope of each item and may vary based on complexity. HST, government fees and other disbursements are extra. Your fee is fixed and agreed in writing before we begin.
Where founders like you usually start
- Step
Put the code in the company.
Incorporate, sign founder IP assignments and vesting, and collect a written assignment from every contractor or agency that touched the code. Paying an invoice doesn’t transfer ownership.
- Step
Write terms for what your AI really does.
AI product terms on inputs, outputs and training, a privacy policy that matches your data flows, and terms of service or a SaaS agreement around them, all saying the same thing.
- Step
Get ready for the first enterprise customer or cheque.
A SaaS agreement that can take a procurement review, and a clean record of ownership and equity before investors start their due diligence.
Brooke & For Founders Law brought sharp legal judgment to every stage of building DMP — from our SAFE deal suite and shareholder agreement to IP assignment, privacy compliance and our terms of service. Her ability to move fast without cutting corners gave us the confidence to build on solid legal ground. We couldn’t have gotten here without her.
Questions
Are our terms of service enough, or do we need AI terms?
AI product terms usually sit alongside your terms of service or SaaS agreement. They say who owns inputs and outputs, whether you train on customer data, that output can be wrong and should be reviewed, and which third-party models process the data. They have to match your privacy policy.
Can we train our models on customer data?
It depends first on your contracts: many customer agreements and data processing terms limit you to processing data on the customer’s behalf. Where the data includes personal information, privacy law also calls for an identified purpose and meaningful consent for new uses. Decide your position and state it plainly in your terms.
Is AI regulated in Canada yet?
Canada has no general AI statute; the proposed Artificial Intelligence and Data Act died in January 2025. AI products are still governed by privacy law, the Competition Act, human rights and employment law, and sector rules. If you have users in the EU, map your product against the EU AI Act too.
Who owns code a contractor or agency wrote for us?
Under the Copyright Act, the contractor does unless they sign a written assignment to your company. Employees are different: the employer is generally the first owner of work made in the course of employment. Moral rights stay with the individual unless they waive them.
Are we responsible for what our chatbot tells customers?
Generally yes. In Moffatt v. Air Canada, a British Columbia tribunal held the airline responsible for wrong information its chatbot gave a customer. Limiting what the bot answers, grounding it in accurate sources and clear terms are the practical steps.
Can we host Canadian customer data in the US?
Generally yes. PIPEDA doesn’t require data to stay in Canada, but you stay responsible for it, must protect it by contract with your providers, and should tell people it may be processed outside Canada. Quebec requires an assessment first, and some customers require Canadian hosting by contract.
General information, not legal advice.
Read next
- Guide What should my AI product terms of service cover?
- Guide What goes in a SaaS customer agreement?
- Guide Who owns code a contractor writes?
- Guide Do I have to tell people my product uses AI?
- Guide Is my company liable for what its AI chatbot tells customers?
- Insight Can your SaaS use customer data to train its AI models?
- Insight Is AI regulated in Canada? What startups need to know in 2026, including the EU AI Act
- Insight Open-source licences in AI products: code, model weights and datasets
Tell us what you’re building on a free 15-minute call. Book a Free Consult