CAWS
Work with CAWS

Bring workers' own evidence into your AI decisions.

CAWS works with public agencies, private employers, and nonprofits, and with unions, worker centers, safety committees, and staff groups. Any of them can bring CAWS in for training on how AI is showing up in their work and what to do about it, or for an independent, structured look at how AI and algorithmic systems are affecting the people doing the job. Engagements are built around the direct experience of the people using these systems, and end in decisions someone can act on.

Start a conversation
Who this is for

If AI directs, monitors, or evaluates the work, this is for you.

CAWS is built for any workplace where AI or algorithmic systems direct, monitor, measure, or evaluate the work. It focuses first on high-stakes work, where the safety of workers and the public depends on getting it right, and where the cost of getting AI wrong is not measured in productivity alone.

Worker

Anyone whose work is directed, monitored, measured, or evaluated by an AI or algorithmic system, whatever their sector, title, or employment arrangement. That includes office, professional, and technical staff as much as frontline and field roles, and people who would not describe themselves as workers at all.

AI and algorithmic systems

The specific systems that make or shape decisions at work. The scheduler that assigns the hours, the score that flags someone for review, the tool that drafts the assessment, the system that ranks the applications, the scanner that sets the pace, the camera that watches the floor. Not all of them are called AI, and the label matters less than what the system decides.

Where CAWS is focused first
Public sector Healthcare Disaster & emergency response Hazardous materials Warehousing & logistics

The framework extends to any workforce. These are some of the sectors where the stakes are highest, and where CAWS is focused first.

Don't see your work here? Tell us where AI is showing up in your work.

How an engagement works

One repeatable cycle, with workers' evidence at the center.

CAWS runs a structured worker-feedback loop that surfaces how AI is actually operating, interprets the risks and opportunities with the workers experiencing them, and turns that evidence into decisions the organization can act on. Individual responses are confidential and reported only in aggregate.

1

Map

Identify which AI and algorithmic systems are present in the workplace and which work decisions they affect. This step routinely surfaces systems leadership did not know were running.

2

Listen

Collect worker experience directly through a structured survey designed to disconfirm as well as confirm. It asks about capability and benefit alongside harm, so the findings reflect what workers actually report.

3

Deepen

Survey results determine what comes next. Interviews and focus groups investigate the issues, roles, and sites workers prioritized, rather than a fixed script.

4

Diagnose

Quantitative and qualitative evidence are analyzed together to identify patterns, variation across roles and sites, risks, useful applications, and capability gaps.

5

Validate

Findings return to workers before they are treated as final. CAWS does not present a finding that workers have not had the chance to review.

6

Decide

Leadership receives specific decisions, options, and tradeoffs, not a stack of findings. Every engagement ends in a decision session.

7

Act & reassess

The organization moves on training, governance measures, bargaining language, reporting processes, or further research. Conditions are measured again, and the loop continues.

Ways to bring CAWS in

From a single briefing to a full assessment.

Engagements range from a short leadership briefing to a complete assessment and the resources that follow. Each one is scoped to the organization and the systems in use. Partners own what CAWS builds for them, and CAWS keeps and openly publishes its methods and anonymized aggregate learning. Pricing is provided on request.

Starting point

Executive Briefing

A focused session for leadership on what AI and algorithmic systems are doing to work, how to read the risks and opportunities, and what a worker-centered response looks like. A common front door.

Focused read

Worker Pulse

A structured survey of the affected workers on a defined question or system, analyzed and returned as findings. A fast, contained way to see what workers are actually experiencing.

The core engagement

Full Cycle Assessment

The complete loop, from mapping the systems in use through a decision session with leadership. Findings are validated with workers before they are final.

Capacity building

Training & train-the-trainer

A four-tier ladder on one method, from a 120-minute awareness session to applied work on a system active in your workplace. Delivered directly or built into the organization through train-the-trainer. See the courses.

Resources

Governance & evaluation

Model contract language, notice and appeal protocols, and worker-defined criteria for assessing AI tools before and during deployment. Built with real use and published as common goods.

CAWS is not affiliated with any vendor and does not promote, sell, or certify any AI product or tool. CAWS sells its own services, and nothing else.

Start a conversation

Tell us what you are seeing.

If AI is already running in your workplace and you want an independent account of what it is doing to the work, get in touch.

Or use the inquiry form on the home page.

CAWS wants to hear from workers directly. If you are a worker and not here on behalf of an organization, write anyway.