Safe AI Setup
Safe AI Setup: lock down how your team uses AI
Banning AI does not work; people just use it quietly, and your data goes with them. Safe AI Setup closes the gaps found in your AI Safety Check, so your team uses only approved tools under clear rules, for teams of up to 50 staff.
What we lock down
- Approved tools on the right business plans, with agreements in place where your work needs them (for example, a Business Associate Agreement for patient information)
- Access controls: single sign-on where available, the right people on the right tools, and leavers removed
- Safe settings: sharing, retention and model-training settings set as protectively as each plan allows
- Logging: a record of what is in use, so you can see it and show it
- A plain AI use policy covering which tools are allowed, what data never goes in, and when a human must review the output
- Staff training and sign-off: everyone trained, and each person attests they have read the policy
Need more than AI? We also handle cybersecurity, cloud and lab equipment security.
Built around your rules
- Life sciences: AI kept away from critical GxP decisions, with human review and documentation near any quality record, in line with the FDA-EMA guiding principles for good AI practice
- Clinics: patient information only in tools covered by a Business Associate Agreement on the plan staff actually use
- Law firms: a policy organized around the duties in ABA Formal Opinion 512
- Accounting and tax: client return data kept out of unapproved tools, with consent questions handled with your counsel
What you have at the end
- An approved-tools list with the reason for each
- A signed AI use policy and staff attestations
- A configuration and access record for each tool
- A handover pack, plus the option to keep it all current with Managed Safe AI
How it fits together
AI Safety Check (10 days). Find every AI tool your team uses, what data goes into it, and the biggest risks. The fee is credited in full to Setup.
Safe AI Setup (up to 50 staff). Every Setup starts with an AI Safety Check. Lock it down: approved tools, access controls, a clear policy, logging, and every staff member trained and signed off.
Managed Safe AI (monthly, 3-month minimum). Keep it safe as tools, vendors, staff and rules change.
Our guarantee: if the agreed controls are not in place by the agreed date, we keep working free until they are.
Why Assets Digital
Founder-led: every engagement is led by Shane Donoher, whose background is in security, infrastructure and regulated technology. You work with the person doing the work.
"Assets Digital has been instrumental in helping us with technical design and implementing our IT strategy, security, and network infrastructure in very short time… Shane and his team are outstanding." — Zachary Betts, Google review
"I didn't realize how many manual tasks we could automate until we started working with Assets Digital. Superb team. Super friendly. Competent and responsive!" — Wade Houston, Google review
"In one session Shane solved them for us… He is a wealth of incredible knowledge which he readily shares so patiently and in 'lay speak.'" — Katherine Rothman, Google review
FAQ
How big a team does Setup cover?
Up to 50 staff. Larger teams are scoped separately.
Can we skip the Safety Check?
No. Setup is scoped from Safety Check findings, which keeps it fixed and predictable, and the Safety Check fee is credited to Setup.
What if the controls are not in place by the agreed date?
We keep working free until the agreed controls are in place.
Who writes the policy?
We draft it in plain English. Your leadership and counsel review and approve it. We do not give legal advice.
Do we have to switch tools?
Not always. Often it means moving staff onto the business plan of a tool they already use, and setting it up properly.
Lock it down
It starts with a 10-day AI Safety Check.
Book an AI Safety CheckNot legal advice. Assets Digital is not a law firm and does not certify compliance. We help you put sensible controls in place; your counsel, compliance or quality lead decides what your rules require.