AI legal counsel for companies building with and buying AI
AI contracts, AI policies and risk allocation for startups, SaaS companies and businesses that build with AI or buy it, from a California attorney who has studied AI since 2016 and has built software with it and on top of it.
Legal services are offered under California and federal law. Not sure if your question fits? Reach out — we'll let you know.
Why Work With an AI Lawyer Who Has Built With AI
Most AI contract problems sit in the gap between what the paper says and what the system does: where prompts are retained, whether outputs can be reused, who is liable when a model gets it wrong. Stuart Wilson has studied that gap since 2016, wrote his law school thesis on AI and liability in 2019, and has built software on top of AI models, so he can tell what a vendor's "we don't train on your data" actually covers and which risks a contract can allocate and which it can't.
The aim is agreements and policies that fit how your product or business really uses AI, not generic terms with "AI" added.
AI Contracts, Policies and Compliance
AI Vendor Agreements and API Terms
Reading the terms before you build on them
Model providers and AI vendors each draw the lines differently. We review what happens to your data and whether it can be used for training, who owns the output, what the indemnities actually cover, and what you are promised about availability, changes and termination.
AI Features in Your Own Product
Terms, disclaimers and who bears the risk
If your product calls an AI model, your customers need terms that match what it does. We draft terms of service, disclaimers and disclosures, and allocate liability between you, your customers and your upstream providers so that your promises are not broader than the ones you receive.
Ownership of AI-Generated Content
Who owns what the model produces
Ownership of AI output is unsettled in places, and contracts and licenses often assume more than the law guarantees. We help you set realistic expectations and write them into your agreements with customers, contractors and vendors.
AI and California Privacy Law
Data flows, notices and California privacy law
When personal information is sent to a model or used to improve one, privacy obligations follow it, including those under California consumer privacy law (CCPA). We map the data flows and align your privacy notices and vendor terms with what really happens.
AI Acceptable Use Policy for Employees
Clear rules your team will actually follow
Employees are already using AI tools. A short, practical policy covers which tools are approved, what information must never be entered, how output is checked, and what is disclosed to customers.
AI in Diligence and Vendor Review
Answering and asking the hard questions
Investors, acquirers and enterprise customers now ask how you use AI and where your code and content came from. We help you prepare accurate answers, and we help you review the AI claims of the vendors you are considering.
AI Legal Help for Startups, SaaS and Buyers of AI Tools
- Software and SaaS companies that call AI models from their products
- Founders and small teams building with AI-assisted development
- Businesses buying AI tools and wanting the contract read before signing
- Companies writing a first AI policy for employees
- Companies answering AI questions from investors, acquirers or customers
How We Use AI in This Practice
The practice may use AI in drafting, research and other work. Stuart reviews all AI-assisted work himself. The practice's policy is not to enter confidential or identifiable client information into commercial AI tools, and client data is not used to train AI models. If you would like to know whether AI was used on your matter, ask and we will tell you. Using AI does not change what you pay.
Take the AI Readiness Check
A short yes/no checklist on how your business uses AI. You get general topics worth a closer look, with related reading.
AI Readiness Check
Companies that build products with AI or buy and use AI tools in their business.
This check gives general information, not legal advice. Every situation is different. Using it does not create an attorney-client relationship, and your answers stay in your browser unless you choose to send them.
8 yes/no/not sure questions, about 2 minutes. No account needed.
Topics worth a closer look
These topics are general pointers, not legal advice or a review of your business. Every situation is different.
-
AI acceptable use policy
Staff may use AI tools on their own, and without written rules, confidential or customer information can end up in tools the company has not reviewed.
Related reading
Related service: AI Acceptable Use Policy ($900)
-
AI policy currency
AI tools and the rules around them can change quickly, so an older policy may not match the tools people actually use.
Related service: AI Policy Review ($500)
-
AI vendor agreements
AI vendor terms can differ widely on whether inputs may be used for training, who owns outputs, and what the vendor promises about its service.
Related reading
Related service: AI Vendor Agreement Review ($650)
-
Customer data and AI
Privacy rules and customer contracts can limit how personal and confidential data may be used, which may include sending it to an AI provider.
Related reading
Related service: AI Vendor Agreement Review ($650)
-
Human review of AI output
AI output can be inaccurate or resemble existing material, so a review step is often used as a safeguard before content reaches customers or the public.
Related reading
Related service: AI Acceptable Use Policy ($900)
-
Ownership of AI-generated work
Ownership and protection for AI-assisted work can be unsettled, and vendor terms may add their own conditions.
Related reading
Related service: AI Vendor Agreement Review ($650)
-
AI product legal review
AI features can raise questions about customer terms, disclosures, data use and responsibility that may be easier to address before launch than after.
Related reading
Related service: AI App Legal Review ($1,200)
-
AI in customer terms and privacy policy
Customers may expect clear disclosure of AI use, and older terms may not mention it.
Related reading
Related service: AI Law and your business
None of these topics were flagged by your answers. This is general information only, not a conclusion about your business. If you have a specific question, a free discovery call is a simple way to talk it through.
Want us to follow up? (optional)
Please do not include confidential details. Sending this does not create an attorney-client relationship. Disclaimer & Privacy Notice
Pick a time for your free discovery call
Your answers suggest a 30-minute discovery call is a good next step. This booking link is just for you and works once.
Trouble seeing the calendar? Open the booking page in a new tab.
A good place to start
For a business at this stage, a Foundation Planning Session ($275) is often the most useful first step: a focused hour on your questions, prepared in advance, with a written summary afterwards.
Book a Foundation Planning SessionAbout existing disputes
Disputes, demand letters and lawsuits are outside what this practice handles, so a call here would not be the right fit. We will reply by email and, where we can, point you toward someone better suited.
AI Legal Questions, Answered
These answers are general information, not legal advice. Every situation is different, and reading them does not create an attorney-client relationship. For advice on your situation, book a discovery call. See the disclaimer and privacy notice.
Do I need a lawyer to review an AI vendor's terms?
You are not required to have one, but AI vendor and API terms often decide who can use your data, who owns the output, and who pays if something goes wrong. Those are the clauses worth having read before you build on a provider or sign an enterprise agreement.
Who owns content created with AI?
It depends on the facts and on the contract. The law on ownership of AI-generated content is still unsettled in places, and a vendor's terms may assign, license or limit your rights in the output. Reviewing those terms and setting realistic expectations in your own agreements is the practical starting point.
Does my company need an AI use policy?
If employees use AI tools, a short written policy helps. It usually covers which tools are approved, what information must never be entered, how output is checked before use, and what is disclosed to customers.
Can an AI vendor train on my data?
Some vendors do by default, some only on certain plans, and some never. The answer is in the vendor's terms and privacy documents, and it can change between consumer and business tiers. Check it before sending confidential or personal information to a tool.
What does a contract involving AI need that a normal one does not?
Usually clear terms on data use and training, ownership of output, who is responsible for errors or infringing output, and what you may promise your own customers about how the AI behaves.
Do you work with companies outside California?
Legal services are offered under California and federal law, and clients can be located anywhere. If your question turns on the law of another state or country, reach out and we will tell you whether it is something we can help with.
How is AI legal work priced?
Four AI services are offered at flat rates: AI Vendor Agreement Review, AI Acceptable Use Policy drafting, AI App Legal Review and AI Policy Review. Each states what it includes and excludes, and the cap and any hourly rate beyond it are stated in writing before work starts. See the AI section of the flat-rate pricing page for current amounts.
Flat-rate AI legal services are listed in the AI section of the flat-rate pricing page.
Further Reading
Tell us what you are building or buying, and we will tell you what to look at first.