Find out where you stand before the rules are final.
Book a free 30-minute call, or send a short brief and Mike will reply with a first read on scope.
Colorado's automated decision-making technology law takes effect January 1, 2027. We find each consequential decision your software touches, list the gaps with an owner for each, and put the notices and review paths in place before that date.
RIG is not a law firm and this is not legal advice. We work alongside your counsel.
An applicant screen inside your hiring software, a risk score inside a lending tool, a ranking inside a tenant portal. Many firms switched these on without calling them AI.
Under the law, developers and deployers both carry duties. A vendor statement speaks to their model, not to how your team uses it.
Counsel can tell you what the law requires. Someone still has to send the notice, take the correction request and route the human review, every time.
What it costs to wait: rules are still being finalized this fall, and the effective date does not move with them. Every month you wait is a month less to inventory, fix and test before January 1, 2027.
| Date | What happened | Source |
|---|---|---|
| 2024 | SB 24-205 signed, creating protections against algorithmic discrimination in consequential decisions made by high-risk AI systems. | leg.colorado.gov |
| May 2026 | SB 26-189 signed. It repeals and reenacts those provisions with requirements for automated decision-making technology used in consequential decisions, for both developers and deployers. | coag.gov/ai, leg.colorado.gov |
| August 11, 2026 | The Department of Law filed proposed rules to implement the law. | coag.gov/ai |
| October 2026 | Interim draft rules released October 6; rulemaking hearing held October 26. | coag.gov/ai |
| January 1, 2027 | The law and its provisions take effect. Consumers can request and correct inaccurate personal data used by these systems. | coag.gov/ai |
Read on October 26, 2026. Rules may change before they are final; check the Attorney General's page for the current text.
Thirty minutes with Mike on where AI touches decisions in your firm and whether the law is likely to reach you.
Free
Every place AI or automated scoring touches a hire, an application, a price or a denial, and what is missing for each.
Scoped after a short brief
Everything in P2, plus written assessments per system and the notice, correction and review paths built into the workflow and tested.
Scoped after a short brief
| What is inside | P1 Readiness call | P2 Inventory and gap list | P3 Audit and controls |
|---|---|---|---|
| Conversation about your systems | Yes | Yes | Yes |
| List of every AI-assisted decision, by system and data source | No | Yes | Yes |
| Each decision marked consequential or administrative, with the reason | No | Yes | Yes |
| Vendor settings that moved a system into scope | No | Yes | Yes |
| Gap list with one named owner per gap | No | Yes | Yes |
| Questions for your counsel, ready to send | No | Yes | Yes |
| Written assessment per covered system | No | No | Yes |
| Consumer notice and correction path in the workflow | No | No | Yes |
| Human review path for adverse decisions, tested | No | No | Yes |
| Record of what changed, who approved it and when | No | No | Yes |
In racing, every car passes inspection before it is allowed out. Same idea here: we check each system before the date, not after the first complaint.
Mike Rodgers runs Rodgers Intelligence Group. He does the inventory with your team, in your tools, and signs his name to the gap list.
Where the work turns into building controls, it follows the same approach as our agent governance work. If you want to start smaller, a One-Workflow Review looks at a single process first.
Best for privileged advice on scope and liability. Check whether it names your actual systems and decisions.
Best for tracking controls and evidence. Check whether it can find a scoring feature switched on inside another vendor's tool.
Best for what their model does. Check whether it says anything about how your team uses it.
Best for finding every decision in your own workflow and wiring the notices and reviews in. This is what RIG does, next to your counsel.
Book a free 30-minute call, or send a short brief and Mike will reply with a first read on scope.
It reaches developers and deployers of automated decision-making technology used in consequential decisions, for example in hiring or lending. Whether it reaches you is a legal question for your counsel; the inventory gives them the facts to answer it.
No. RIG is not a law firm. We find and document how decisions are made and build the operating controls. Your counsel advises on the law.
The effective date is January 1, 2027. The inventory does not depend on the final rule text, and it is the longest part of the work.
Colorado also passed a Chatbot Safety Act with the same effective date. If you run a customer-facing chatbot, we flag it in the inventory for your counsel.
Yes. We review systems and settings inside your access rules. If you need AI that never leaves your firm, see Private AI.
Sources: Colorado Attorney General, Artificial Intelligence; SB 26-189; SB 24-205.