California Is Building a Paper Trail for AI Workplace Decisions

You may not think of your professional services firm as a tech employer. But if you use AI to schedule client assignments, flag performance issues, screen candidate resumes, or help decide who gets promoted — and if any of your employees live in California — two advancing California bills would create a new documentation obligation you don't currently have.

AB 1898 and AB 1883 are the most broadly scoped AI employment bills advancing in the US in 2026. Illinois, New York City, and Colorado have already passed targeted laws covering specific uses of AI in hiring. California's approach is different: it covers ongoing AI-assisted workplace decisions, not just initial hiring — and it applies to any employer of California residents, regardless of where the employer is headquartered.

What the Bills Would Actually Require

AB 1898 targets automated employment decision tools. Under the bill, before or at the time an employer uses an AI system in a covered employment decision — hiring, firing, promotion, demotion, suspension, or performance evaluation — the employer must provide written notice to the affected employee that an AI tool was used in that decision.

The notice would need to identify the AI tool, describe its role in the decision (screening, scoring, ranking, or final determination), and provide contact information for a person at the employer who can answer questions about the AI system's use.

AB 1883 covers AI-assisted workplace monitoring and productivity tracking. If an employer uses AI to analyze work patterns, productivity levels, communication data, or location information for the purpose of employment decisions or performance management, written notice is required.

Together, the bills cover the full AI-assisted employment workflow: initial candidate screening, ongoing performance tracking, and final employment decisions.

Why Professional Services Firms Are Exposed

You may not think of yourself as "using AI in employment decisions." But consider the workflows where AI is now embedded in professional services practices:

Scheduling and workload allocation: Several matter management and staffing platforms now use AI to suggest which attorney, accountant, or consultant should be assigned to a client. If that suggestion influences who gets premium client access, and if that access influences compensation or promotion, California regulators would likely characterize it as an AI-assisted employment decision.

Performance review inputs: AI-assisted billing analysis, productivity dashboards, and client satisfaction tools increasingly surface data that feeds into performance reviews. If an AI tool flags that an associate is billing 20% below their peer group and that flag appears in a review conversation, the AI may have participated in the evaluation.

Resume screening for internal hires: Promoting internal staff to new roles, recruiting for lateral hires, and screening resumes for open positions — all common uses of AI-assisted tools — fall within AB 1898's scope for covered employment decisions.

Scheduling AI for hybrid work management: If you use AI to optimize scheduling for office days, remote work, or client-facing coverage, and if those scheduling decisions affect compensation, visibility, or advancement, AB 1883's monitoring provision may apply.

The issue is not whether you consciously "made an employment decision with AI." It's whether AI was present in any process that influenced employment outcomes.

The Documentation Burden

The practical compliance requirement is a disclosure process. Before or contemporaneous with any covered employment decision, a written notice must be provided to the affected employee.

For a firm with 10-20 employees in California, that means:

1. Inventory: Which AI tools are used in your practice, and which of those tools touch processes that affect California employees?
2. Mapping: For each AI-touched process, what employment outcomes does it connect to?
3. Triggering: At what point in the process is notice required — before the AI analysis runs, or before the decision is communicated to the employee?
4. Delivery: How is the written notice provided — email, HR system record, employment document?

This is less complex than it sounds for firms with a small California headcount. For a 15-person firm with 2 California-based associates, an hour of process mapping and a notice template is the whole compliance program.

For a 40-person firm with 8 California employees and AI embedded across scheduling, billing analysis, and performance review, it's more involved — but still operationally manageable if you build it before the law passes rather than scrambling after.

The Multi-State Compliance Pattern

California is not acting alone. The Transparency Coalition's March 6, 2026 AI legislative update tracked 17 bills across 11 states that address AI in employment decisions. Colorado's CPAIA (live June 30, 2026) covers AI systems used in consequential decisions, including employment. Illinois HB 3773 (live January 1, 2026) requires notice when AI is used in employment decisions. New York City Local Law 144 (live since 2023) requires bias audits for automated employment decision tools.

The pattern: each state is building a notice, documentation, or audit requirement for AI-assisted employment decisions. The requirements differ in scope and specificity, but the direction is uniform.

For a professional services firm with staff across multiple states, a single AI use policy and a standardized disclosure template is more efficient than tracking each state's specific requirements individually. The policy should:

• Name the AI tools used in your practice
• Identify which employment processes those tools touch
• State the firm's standard of human review for AI-assisted decisions
• Describe how employees can ask questions about AI's role in decisions affecting them

This policy — one page, written in plain language — is the foundation that addresses California's requirements and provides a starting point for every other state with active AI employment legislation.

What to Do Before the Bills Pass

AB 1898 and AB 1883 are advancing through committee as of March 2026. If both pass and are signed, expect a 6-12 month implementation period before enforcement begins.

That timeline is not a reason to wait. The firms that will struggle are the ones that haven't mapped their AI use to their employment workflows before implementation arrives. The firms that will comply smoothly are the ones that built that map now, when there's no pressure.

This week:

1. List the AI tools your firm uses that touch any part of your employee or hiring workflow. Scheduling tools, billing analysis dashboards, performance management platforms, ATS systems.
2. For each tool, note whether any of its outputs are used in employment decisions — even informally.
3. If you have California employees, add a calendar reminder to check AB 1898 and AB 1883 status in 60 days. Track at the Transparency Coalition's legislative tracker (transparencycoalition.ai) or at leginfo.legislature.ca.gov.

You don't need an employment lawyer to do step 1 and 2. You need an honest inventory of how AI is embedded in your people processes. The inventory is also useful regardless of California law — it answers the governance question that your own clients are starting to ask.