Colorado AI Act & Employment: Non-Competes, Hiring AI, and SB 205 Compliance

Colorado employers are navigating a uniquely complex regulatory environment. Two significant laws intersect in ways that many HR departments and employment attorneys haven't fully reckoned with:

1. HB 22-1317 (effective August 2022) — Colorado's strict non-compete restriction, which limits non-compete agreements to workers earning above $123,750 (adjusted annually) and requires specific notice and consideration provisions.
2. SB 24-205 (effective June 30, 2026) — The Colorado AI Act, which requires comprehensive compliance programs for any AI system that makes or influences "consequential decisions" — and employment decisions are explicitly listed as consequential.

The connection might not be obvious, but it matters: if you're using AI tools for hiring, promotion, compensation, performance evaluation, or termination — the same decisions that intersect with non-compete enforcement — you now have AI compliance obligations layered on top of your existing employment law duties.

This isn't theoretical. The Colorado AG has signaled that employment is a priority enforcement area for SB 205, precisely because AI hiring tools have the highest potential for algorithmic discrimination across protected classes.

Related: complete SB 24-205 guide · AI hiring tools compliance guide · bias audit process

Before diving into the AI intersection, here's where non-competes stand in Colorado as of 2025:

• Highly restricted. HB 22-1317 made Colorado one of the most restrictive states for non-compete agreements. Non-competes are generally void unless the worker earns above a salary threshold (adjusted annually by the Colorado Department of Labor).
• 2025 threshold: The adjusted earnings threshold for enforceable non-competes is approximately $123,750+ (exact figure confirmed annually). Below this amount, non-competes are unenforceable.
• Notice required. Employers must provide clear written notice of any non-compete before the worker's start date or, for existing employees, at least 14 days before the agreement takes effect.
• Penalties. Employers who attempt to enforce void non-competes face penalties of up to $5,000 per violation, plus actual damages to the worker.
• Non-solicitation carve-out. Customer non-solicitation agreements are permitted for workers above a lower threshold (roughly $61,875 in 2025), with similar notice requirements.

Now here's where AI enters the picture: many employers use AI-powered analytics to identify "flight risk" employees, predict which workers are likely to leave, and flag individuals for non-compete enforcement. Under SB 205, these AI systems are high-risk because they influence employment-related consequential decisions.

If you use any of the following AI tools, they're almost certainly "high-risk AI systems" under SB 24-205:

• Resume screening / ATS ranking — HireVue, Greenhouse AI, Workday AI, Lever, iCIMS
• Video interview analysis — AI-powered facial expression or speech pattern analysis
• Skill assessments — AI-scored testing platforms, coding assessments with AI grading
• Performance evaluation — AI-generated performance scores, stack ranking algorithms
• Workforce planning — AI-driven headcount optimization, layoff modeling, attrition prediction
• Compensation analysis — AI tools that recommend salary ranges or flag pay equity issues
• Flight risk prediction — Any AI system that predicts which employees are likely to leave

For each of these systems, SB 205 requires:

1. Registration in your AI system inventory with documented risk classification
2. Annual impact assessment covering purpose, data inputs, bias risks, and human oversight
3. Regular bias auditing across race, gender, age, disability, national origin, sexual orientation
4. Consumer disclosure — job candidates and employees must be informed when AI influences decisions about them
5. Incident response — if bias is detected, you have 90 days to notify the AG and affected individuals

The disclosure requirement alone is a sea change. Imagine telling every job applicant: "An AI system will evaluate your resume and rank you against other candidates." Many employers aren't prepared for that conversation — but the law requires it.

Colorado employers now face a double compliance obligation that no other state has created:

• Non-compete compliance (HB 22-1317): Ensure agreements meet salary thresholds, provide proper notice, don't enforce void agreements
• AI compliance (SB 24-205): Ensure any AI tools used in the employment lifecycle are documented, assessed, monitored, and disclosed

The risks compound. An employer who uses AI to identify "flight risk" employees and then selectively enforces non-competes based on AI predictions is:

1. Using a high-risk AI system (flight risk prediction) without required compliance
2. Potentially engaging in algorithmic discrimination if the AI model correlates with protected class characteristics
3. Possibly enforcing non-competes in violation of HB 22-1317 if the AI doesn't account for salary thresholds

This is a litigation triple threat. An employee who is targeted for non-compete enforcement based on biased AI predictions could have claims under both statutes — and the AG could pursue enforcement on the SB 205 side independently.

Here's a practical compliance roadmap that addresses both obligations:

1. Separate your AI compliance from your non-compete compliance — but coordinate them. Different teams may manage each, but they need to communicate. If your AI tools feed into non-compete decisions, that's a high-risk intersection that needs special attention.
2. Inventory every AI tool in your HR/people stack. Include ATS, HRIS, performance management, learning management, and workforce analytics platforms. If it has "AI" or "smart" features, document it.
3. Assess bias risk in hiring AI specifically. Employment AI has the highest risk of disparate impact across protected classes. Run bias audits before the deadline — finding and fixing issues now is infinitely better than discovering them during an AG investigation.
4. Build disclosure language now. Draft the notices candidates and employees will receive when AI influences decisions about them. Have employment counsel review them alongside your non-compete language.
5. Document everything. The affirmative defense requires proof of good-faith compliance. CO-AIMS creates a continuous, timestamped audit trail — every assessment, every audit, every policy update — that serves as your compliance evidence bundle.

The Colorado employment landscape is evolving rapidly. The intersection of AI regulation and employment law will only get more complex. Organizations that build robust compliance infrastructure now will be positioned not just to survive enforcement, but to turn compliance into competitive advantage — demonstrating to employees, candidates, and clients that they take AI ethics and fairness seriously.