Course Overview & Syllabus
Why Seattle is the sharpest case study in the country for algorithmic policing governance, and how this course is organized around the five questions that actually determine whether a tool gets deployed.
Who This Course Is For
This course is written for police department leadership, policy staff, and oversight personnel who sit on the approving side of algorithmic policing decisions — the people who have to answer, in public, why a tool was acquired, how it was reviewed, what it does operationally, whether it works, and who watches it once it’s live.
That is a different job than using the tools day-to-day. This course does not teach officers how to operate a specific product. It teaches command staff and policy leads how to evaluate, justify, and govern the whole category — using Seattle as the working example, because Seattle has more publicly documented history on this than almost any other city in the country.
Why Seattle Is the Case Study
Three things collided in Seattle over the last decade, and the collision is why this city’s paper trail is so useful to study:
- An unusually aggressive local ordinance regime. Seattle passed one of the first municipal surveillance-technology ordinances in the nation (2013), then rewrote it into a much stronger disclosure-and-approval regime (SMC 14.18, 2017) — years before most cities had any equivalent rule.
- A federal consent decree that shaped department culture for over a decade. Seattle Police operated under direct federal oversight for the use of force and bias-free policing, which means every new technology decision was made inside an environment already primed for public scrutiny.
- Rapid, real deployment of AI tools regardless of the above. Real-time crime centers, bias-crime NLP classifiers, AI-assisted report review, and predictive/forecasting models have all been piloted or deployed in Seattle — sometimes ahead of the policy written to govern them, which is itself an important lesson.
That combination — strong ordinance, exiting federal oversight, live AI deployment — is what makes Seattle the sharpest available teaching case for how algorithmic policing governance actually plays out, warts and all.
How This Course Is Organized
The course follows five modules, each built around one governing question:
| Module | Question | What You’ll Be Able to Do |
|---|---|---|
| 1. Regulatory Framework | What law and policy actually govern this? | Trace Seattle’s ordinance history and explain what a Surveillance Impact Report requires |
| 2. Technical Implementations | What is actually deployed, and how does it work? | Distinguish operational, accountability, tactile, and integrated AI systems by function |
| 3. Impact Analysis | Does it work, and at what cost? | Weigh documented operational gains against systemic drawbacks, using ShotSpotter as the central case |
| 4. Criminological Research | What does the independent evidence say? | Cite specific findings on AI report quality and predictive forecasting models |
| 5. Oversight and Reform | Who watches it now, and what should change? | Explain the post-consent-decree oversight structure and evaluate concrete policy recommendations |
Each module builds on the last. Module 1 gives you the legal vocabulary and process everything else refers back to; by Module 5 you’ll be applying it to make an actual recommendation.
Course Objectives
By the end of this course, you will be able to:
- Explain Seattle’s surveillance ordinance regime — Ordinance 124142 (2013), SMC 14.18 (2017), and the 2025 AI policy layered on top — and what triggers council review.
- Name and categorize the AI systems currently in use or piloted by Seattle PD, sorted by what job they actually do (operational, accountability, tactile, integrated).
- State the documented operational advantages and systemic drawbacks of algorithmic policing tools, with specific figures rather than vendor talking points.
- Summarize the independent research on AI-assisted report quality and predictive forecasting accuracy, including where it contradicts vendor claims.
- Describe the current oversight structure — post-consent-decree — and evaluate the strategic recommendations on the table for closing remaining gaps.
It would be easy to teach this material as either “AI is transforming policing for the better” or “AI in policing is a civil-liberties disaster.” Neither framing survives contact with Seattle’s actual record. Some tools show real operational value with contained risk (bias crime NLP classification, VR de-escalation training). Others show minimal or negative value despite heavy investment (ShotSpotter, some report-writing tools). This course teaches you to tell the difference on a tool-by-tool basis — that discrimination is the actual skill leadership needs, not a category-wide verdict.
Before Module 2, list every AI or algorithmic tool your own department currently uses or is evaluating. For each one, note:
- Which of the four technical categories from Module 2 it falls into (operational, accountability, tactile, integrated).
- Whether it has been through a documented review process equivalent to Seattle’s Surveillance Impact Report.
- Whether you have independent evidence — not vendor claims — that it does what it’s procured to do.
Keep this list. Module 5 will ask you to apply the strategic recommendations framework to it.
Up Next
Module 1 continues with the legal foundation: why cities regulate surveillance technology locally at all, how Seattle’s ordinance evolved across three distinct legislative moments, and what a tool actually has to pass through before it can be deployed — starting with Privacy Localism.