01
Advisory
Decide what AI should do here.
Where you stand, what's worth doing first, what to buy versus build, and how to explain it to a board or a room of trustees.
How it worksAI advisory · Policy · Privacy · Implementation · Agents
AI that works for your organization — and answers to it.
Solidare AI helps unions and funds, nonprofits and public bodies, and growing businesses decide what AI should do, write the rules, protect the data, and build the assistants and agents that do the work. Then we stay to keep it right.
YOUR DATA STAYS YOURS · PEOPLE DECIDE · YOU OWN WHAT WE BUILD
How governed AI runs here
Where organizations are
Your people are already using AI. The question is whether it's on your terms.
Most organizations we meet have no written rules, no map of what data is going where, and a board that hasn't been asked. That's normal. It's also the cheapest moment to get it right.
1 in 5
organizations has a mature governance model for AI agents.
Deloitte, State of AI in the Enterprise 2026
55%
of employers using AI name data privacy as a top worry; 57% name hallucinations.
Gallagher AI Adoption & Risk Survey, Feb 2026
Jan 1, 2027
Colorado and California rules on AI in consequential decisions take effect. Illinois, Texas and California employment rules are already in force.
SB 26-189 · CPPA ADMT regulations
What we do
Five things, in the order most organizations need them.
Start anywhere. Each stands on its own, and each is built to make the next one easier.
02
Policy & Governance
Write the rules people will actually follow.
An AI use policy, an approval path for new tools, oversight roles, and a review cadence — built on NIST AI RMF and ISO/IEC 42001 structure, written in plain language.
How it works03
Privacy & Security
Keep member, employee and client data where it belongs.
Find where sensitive data is already flowing into AI tools, review the vendors handling it, and put controls in place before it becomes an incident.
How it works04
Implementation & Integration
Put AI to work inside the systems you already run.
Private AI in your environment, connected to your records, membership system, LMS, accounting and email — rolled out with training and measured against real hours.
How it works05
Agentic AI
Build agents that do real work, with a person in the loop.
Scoped agents for intake, reconciliation, drafting, scheduling and answering questions from your own material — with approval gates, logs and an off switch.
How it worksHow an engagement goes
Listen. Map. Decide. Build. Stay.
The same path every time, so you always know what's next and what it will cost before we start it.
01
Listen
We sit with the people who do the work and learn how it runs today, including what's already being pasted into chatbots.
02
Map
What data you hold, who touches it, what the law expects of you, and where AI would take real hours off the table.
03
Decide
Policy, priorities, private or public. Written so a non-technical board or a room of trustees can approve it in one meeting.
04
Build
Assistants, connections, agents, in your environment, with approval gates, tested against the threats that matter.
05
Stay
We monitor, tune, and update the rules as AI and the law change. You own everything we build.
Packages
Scoped, timed, and written down before we start.
Six ways to begin. Each has a fixed scope, a timeline, and a list of what you'll hold at the end.
012–3 weeks
AI Readiness Review
Any organization that hasn't yet decided what AI is for here.
What you get023 weeks
AI Use Policy Sprint
Organizations whose staff are already using AI without written rules.
What you get033–4 weeks
Privacy & Security Review
Anyone holding member, employee, patient, donor or client data.
What you get044–6 weeks
Implementation Blueprint
Organizations ready to move from trying AI to running it.
What you get056–8 weeks
Agent Pilot
Teams with one repetitive workflow they'd hand off tomorrow if they could trust it.
What you get06Ongoing
Managed AI Partnership
We keep what's running healthy, keep the rules current as AI and the law change, and sit in as your technology voice when decisions come up.
Industries
Different records, different rules. One standard.
We came up through the union world. The standard we learned there — the data belongs to the people it names — applies to every organization we work with. Who we serve
Business managers, agents, office staff and executive boards.
Trade-union locals
You hold standing, hours, dispatch order, grievances and discipline for every member. AI is already in the office; the question is whether it is on the local's terms.
Where AI helps first →Training directors, coordinators, instructors and trustees.
Training funds & JATCs
Your curriculum is years of instructor work and it is your property. The Department of Labor is bringing AI into registered apprenticeship; funders will ask for a plan.
Where AI helps first →Administrators, trustees, plan professionals and participant-services staff.
Benefit funds & fund offices
Participant data, claims and eligibility carry fiduciary duty and, for health plans, HIPAA. A board needs a record that the questions were asked before AI touched any of it.
Where AI helps first →Executive directors, program leads, development and communications staff.
Nonprofits & associations
Small staff, big responsibility: donor, client and program data, grant reporting that has to be right, and a board that needs to approve what it doesn't fully understand.
Where AI helps first →Agencies, special districts, school and college administrators, elected boards.
Public bodies & districts
Resident data, public-records obligations, procurement rules and open meetings: AI in a public body has to be explainable to the public that funds it.
Where AI helps first →Owners, estimators, project managers and office managers.
Contractors & field services
Estimates, SOPs, customer history and crew records are your edge. Your people are already using AI on them; the question is whether it is inside the law and inside your control.
Where AI helps first →Owners and administrators of law, accounting, engineering and healthcare practices.
Professional firms & clinics
Client and patient confidentiality is the business. AI can take real hours off the table, inside a boundary you can describe to a client, a regulator or a licensing board.
Where AI helps first →Private AI
Your own AI, in your own environment. Nothing leaves.
A model that knows your manuals, your contracts and your records, runs on hardware you own or a tenant that's only yours, logs everything it does, and never trains on your data.
Tier 1 · Local
Runs on hardware you own.
Nothing leaves. Fixed cost. Best for anything that names a person.
Tier 2 · Private tenant
A cloud account that's only yours.
Isolated; model and documents inside it.
Tier 3 · Enterprise chatbot
A vendor's tool, no-training contract.
Fine for general drafting. Data still sits with the vendor.
Engagement examples
What the work looks like, week by week.
Composite examples drawn from our work, details changed, no client named. All six examples →
Training center · AI Use Policy Sprint · 3 weeks
From "everyone’s pasting it into chatbots" to a policy people read
Instructors and office staff at a trades training center were uploading course material and apprentice information into public AI tools. Nobody had written rules, and the training director had been asked by trustees what the plan was.
What they hold now → A policy staff can recite, a record trustees can show, and a 90-day check on the calendar.
Benefit fund office · Privacy & Security Review · 4 weeks
A vendor’s participant assistant, examined before the vote
A fund office was about to sign for a participant-facing chat assistant. The demo was good. The contract said nothing about training, retention or protected health information.
What they hold now → A contract with the answers in it, a resolution in the minutes, and a standing questionnaire for the next vendor.
Solidarity principles
What we will and won't do with AI.
Your data stays yours
Nothing you share with us, or with the AI we build, trains anyone else's model.
People decide
AI drafts, sorts and flags. A person approves anything that touches a member, an employee, a patient or a dollar.
Say what it does
Every AI use is disclosed to the people it affects, in plain language.
Built to be left
You own the policies, the agents and the setup. If you ever want to walk away, you take it with you.
No sales deck
We'll tell you when the answer is “not yet” or “not this.”
What we work to
Frameworks we build on and laws we keep track of.
Aligned, not "certified." We design to these so your program holds up to a regulator, an auditor or a trustee, and so certification is reachable later if you want it.
NIST AI RMF 1.0
The default US vocabulary for AI risk: govern, map, measure, manage. A safe harbor under Texas law.
NIST AI 600-1
The generative-AI profile: twelve risk categories that shape our use policies.
ISO/IEC 42001
The AI management-system standard. We design to its structure so certification is reachable later.
OWASP Top 10 for LLM & Agentic Apps
The threat lists we build against, for assistants and for agents.
MITRE ATLAS
Adversary tactics against AI systems; our threat models map to it.
FPF Best Practices for AI in Hiring & Employment
Industry-endorsed guidance (Aug 2026) we follow for anything touching workers.
Insights
Notes written to be useful.
Law · September 1, 2026
What changes on January 1, 2027
Colorado and California both switch on rules for AI in consequential decisions. If AI touches hiring, discipline, benefits or eligibility, here is what you owe people, and a 90-day way to get ready.
Read the note · 3 minPrivacy · August 18, 2026
Five ways member data leaks into public chatbots
It is rarely a hack. It is a helpful person with a deadline and a free tool. Here are the five routes we find in almost every organization, and the fix for each.
Read the note · 3 minAgentic AI · August 4, 2026
Assistants versus agents: what a director needs to know
An assistant answers. An agent acts. The difference is who is accountable for what happens next, and how much you let it do on its own.
Read the note · 3 minResources
Templates you can use today. No form to fill in.
The questionnaires and templates we hand to clients, as PDFs.
All resourcesQuestionnaire
The ten vendor questions — Send these in writing before the second call. With a scoring column and the three additions for unions and funds.
Template
One-page AI use policy: the structure — The eight sections, the desk card, and what sits behind them. Fill it in with your approved tools and your names.
Board template
Six fiduciary questions and a board resolution — The six questions for the minutes and a one-page AI resolution a board can adopt as written.
Next step
Let's have a working conversation.
Thirty minutes, the people who do the work, and a plain answer about what AI should and shouldn't do here.
NO SALES DECK · NO OBLIGATION · ASK@SOLIDARE.AI