Regulation Guides

Portland Face Recognition Ban: 2026 Compliance Guide

November 21, 2026 · 13 min read

By AICompliant Research Team

Portland Private Face Recognition Ban: What Companies Need to Know in 2026

Most facial recognition compliance work happens at the state level, under biometric privacy statutes that regulate consent and notice. Portland, Oregon took a different path. Its private face recognition ban, codified at City Code ch. 34.10 and enacted through Ordinance 190114, does not ask businesses to collect better consent. It tells them not to use face recognition technologies in places open to the public at all.

For compliance officers and general counsel, that changes the question. Instead of "How do we disclose this properly?", the question becomes "Where is this technology running, and should it be running there?" The answer depends on inventory work that many organizations have not done, especially where vendors embed face matching inside cameras, access systems, or analytics tools. That is a core use case for any AI compliance platform: finding the AI that is already in your operations.

This guide covers the ordinance's scope, its single core requirement, the private-action penalty structure, its effective date, and a practical preparation plan. It also places Portland alongside the other biometric and AI laws your legal team is probably tracking.

Effective Date and Status

Portland's private face recognition ban was adopted as Ordinance 190114 and has been in effect since January 1, 2021. It is not a pending or proposed rule. If you operate publicly accessible locations in Portland, you are already subject to it, and any exposure is accruing now.

You can review a summary of the ordinance in our Portland private face recognition ban regulation page.

Scope: Which Places and Which Businesses Are Covered

The core scope concept is places open to the public. In our regulatory catalog, the ordinance is described as covering stores, restaurants, hotels, and venues. The practical test is whether members of the public can walk in, whether they pay or not.

Examples of settings that warrant a close look:

  • Retail stores using camera-based systems for loss prevention or customer analytics
  • Restaurants, bars, and hotels using cameras for guest recognition or security
  • Entertainment and event venues using entry or crowd-management systems
  • Any location where a third-party vendor provides cameras or software that includes face matching

Two points trip up compliance teams.

The technology may not be labeled as face recognition. A "smart camera" or "video analytics" platform may include a face matching feature. Your contract may describe it generically, so your team has to ask the vendor directly.

The ordinance concerns use, not ownership. A vendor may operate the software, but if it runs in your publicly accessible space, your organization is the one with exposure. Treat your vendors' product configurations as part of your own compliance perimeter.

The ordinance has its own definitions and may include exceptions. Because the summary we work from lists a single core requirement, confirm the current text of City Code ch. 34.10 with Oregon counsel before relying on any exception.

Core Requirement: No Face Recognition in Places Open to the Public

The requirement is short and categorical:

Do not use face recognition technologies in places open to the public (stores, restaurants, hotels, venues).

There are no notice-and-consent steps to follow, no published-policy track, and no opt-out workflow that substitutes for compliance. The control you need is operational: the technology must not be in use in covered locations.

That has an important consequence for how you document compliance. A privacy notice or consent banner will not cure a deployment that the ordinance prohibits. What demonstrates compliance is evidence that the capability is absent or disabled, such as vendor configuration records, written vendor confirmations, and internal sign-off on the location inventory.

Penalties: A Private Right of Action With Daily Exposure

The ordinance is enforced through a private action rather than a per-violation city fine schedule. Under City Code §34.10.050, a person can recover:

  • Damages sustained, or $1,000 per day for each day of violation, whichever is greater, and
  • Other appropriate remedies, and
  • Attorney fees for a prevailing plaintiff who made a written demand at least 30 days before suing.

Why the daily structure matters

The "per day" measure makes duration the main driver of exposure. A system that runs for a quarter without anyone noticing it carries a minimum statutory exposure on the order of $90,000 at $1,000 per day. That is a simple illustration, not a prediction of what a court would award, and the "whichever is greater" language means actual damages could exceed it.

The written demand and fee-shifting

Fee recovery is tied to a written demand made at least 30 days before suit. For defendants, that has two practical implications:

  1. A demand letter is a signal to act quickly. Any capability that should not be running needs to be located and shut off, with a documented record of when and how.
  2. Your legal intake process should route letters that mention face recognition or biometric technology to counsel the same day. Mail room or customer service delays eat into the 30-day window.

Because the daily exposure runs from the first day of violation, the discovery date is not the date that matters. Fast shutdown limits further accrual, but it does not erase days already elapsed.

Where Portland Fits Among Biometric and Facial Recognition Laws

Portland's ban sits alongside a patchwork of other biometric and facial recognition rules. The differences are worth understanding, because a national retailer or hospitality group needs to reconcile them in one program.

| Jurisdiction | Approach | Key duty | Penalty (per our catalog) | |---|---|---|---| | Portland (City Code ch. 34.10) | Prohibition | No face recognition in places open to the public | Greater of damages or $1,000 per day; fees after 30-day written demand | | Illinois BIPA (740 ILCS 14) | Consent regime | Written policy, notice and written release before collecting biometrics | Up to $5,000 per violation | | Texas Bus. & Com. Code s. 503.001 (CUBI) | Consent regime | Notice and consent before capturing biometric identifiers for a commercial purpose | Up to $25,000 per violation | | Washington RCW 19.375 | Consent regime | Notice and consent before enrolling biometric identifiers | Attorney general enforcement under ch. 19.86 RCW | | NYC Local Law 3 of 2021 | Disclosure and sale ban | Post a sign at every customer entrance if collecting or using biometric identifiers; never sell, lease, trade or otherwise profit from the information | Up to $5,000 per violation | | Maryland Lab. & Empl. §3-717 | Employment-specific | Facial recognition in job interviews only with a signed applicant waiver | Not stated | | Florida SB 262 | Opt-out rights | Honor opt-outs of profiling and of voice or facial-recognition data | Not re-read in our catalog |

The comparison points to a design decision. Where one jurisdiction allows biometric collection with consent and another prohibits face recognition in public places outright, a single national deployment cannot comply everywhere. Many organizations respond by setting a location-level rule, such as "face recognition disabled in all Portland public-facing sites." That is easier to enforce than a patchwork of consent flows.

The same biometric systems can also touch other regimes. If a facial analysis system is used in hiring, for example, it may implicate the Illinois AI Video Interview Act (notice, explanation and consent before AI video analysis) or the Maryland waiver rule above.

Portland Is Not the Only Portland Rule

Portland also has a separate ordinance on rent-setting software. Portland prohibited anti-competitive rental practices (PCC 30.01.088, Ordinance 192122) took effect February 17, 2026. It bars landlords and vendors from using, paying for, selling, licensing or providing rent-recommendation software built on competitors' non-public data, with penalties of up to $1,000 per violation. It is unrelated to face recognition, but if you operate in Portland across business lines, your city-level inventory should capture both.

Where Portland Fits in the Wider AI Compliance Landscape

A face recognition ban is one narrow rule in a much larger body of AI law. Most organizations with Portland exposure also face broader obligations, and the same inventory work serves all of them.

Colorado. Colorado SB 26-189 (formerly SB 24-205) applies from January 1, 2027. For teams working on Colorado AI Act compliance, the Colorado SB 26-189 requirements in our catalog include notice before using covered automated decision-making technology in consequential decisions, adverse-outcome disclosure within 30 days with correction and human review for deployers, and documentation, update notices and three-year records tied to developer and deployer roles. Penalties arise under the Colorado Consumer Protection Act, and the amount is not stated in Part 17. See Colorado AI Act.

EU AI Act. The EU AI Act (Regulation (EU) 2024/1689) has been in effect since August 1, 2024. For EU AI Act compliance planning, the high-risk obligations apply from 2 December 2027 (Annex III) and 2 August 2028 (Annex I) under Regulation (EU) 2026/1744. Prohibited practices under Art. 5 carry penalties of up to €35,000,000 or 7% of worldwide annual turnover, whichever is higher. Deployers also face disclosure duties for emotion recognition and biometric categorisation under Art. 50(3)-(5). See EU AI Act.

NYC Local Law 144. For hiring tools, NYC Local Law 144 compliance means candidate and employee notice, publishing the bias-audit summary and tool distribution date before use, and making the data type, source and retention policy available on the website or within 30 days of a written request. Penalties are up to $1,500 per violation, per day. See NYC AEDT Law.

The shared lesson is that every one of these laws starts with the same question: which AI and biometric systems do we actually run, where, and who owns them? A face recognition ban is simply the most blunt version of that question.

How to Prepare: A Practical Plan

1. Inventory every location and system

List every Portland site that is open to the public, then every camera, access, analytics and loyalty system in each. Include systems owned by landlords, security contractors and marketing vendors. The goal is a complete record of whether any system includes face recognition capability, and whether it is enabled.

2. Ask vendors direct questions

Ask each vendor in writing:

  • Does the product perform face recognition or facial matching in any mode?
  • Is it enabled by default, or only by configuration?
  • Can it be disabled by location?
  • What logs show when it was enabled or disabled?

Keep the written responses. Vague answers are a risk signal in themselves.

3. Disable or remove, and document it

Where face recognition is running in a covered location, shut it down and record the date, the person responsible, and the technical evidence. Given the per-day damages structure, speed matters more than perfect documentation, but do both.

4. Update contracts

Add contract terms requiring vendors to disclose face recognition features, to keep them off in Portland public-facing sites, and to notify you before enabling new capabilities in product updates. Software updates are a common route by which a capability appears without a procurement decision.

5. Build a demand-letter protocol

Because fee recovery depends on a written demand at least 30 days before suit, set up a routing rule so that any letter referencing face recognition or biometrics reaches counsel immediately. Pre-assign who locates systems, who contacts the vendor, and who records the shutdown.

6. Reconcile with your other biometric obligations

If you also operate elsewhere, align Portland's prohibition with consent-based obligations under Illinois, Texas and Washington law, and the signage and no-sale rules in New York City. A location-based rule set, with Portland marked as "disabled," simplifies this.

7. Reassess on a schedule

Capabilities change through product updates, acquisitions and new vendors. Reassess periodically and whenever a new site or system is added.

How an AI Compliance Platform Supports Portland Readiness

Manual spreadsheets work for a handful of locations. They break down when you have dozens of sites, several vendors and constant product changes. This is where AI compliance software earns its place: not by making legal judgments, but by organizing the work.

A purpose-built AI compliance platform such as AICompliant helps teams keep a system inventory, map each system to the laws that apply to its jurisdiction, and draft compliance documents for your counsel's review. For a Portland-type prohibition, the useful outputs are a location-by-location inventory, vendor questionnaires, and a record showing what was disabled and when. Automated AI compliance workflows can also flag when a new system or site is added, so the question gets asked before launch rather than after a demand letter.

Teams that want a quick first read on which laws touch their footprint can start with the compliance checker. Ongoing tracking and document drafting live in the dashboard. Tools do not replace legal advice, and every output should be reviewed by qualified counsel before use.

Key Takeaways

  • Portland's ban under City Code ch. 34.10 (Ordinance 190114) has applied since January 1, 2021.
  • The requirement is a prohibition on face recognition technologies in places open to the public, such as stores, restaurants, hotels and venues. Consent and notice are not the framework here.
  • Enforcement is through a private action: damages sustained or $1,000 per day for each day of violation, whichever is greater, plus other appropriate remedies, with attorney fees available after a written demand made at least 30 days before suit (§34.10.050).
  • Inventory, vendor controls and documented shutdowns are the core of a defensible program.
  • The same inventory supports Colorado SB 26-189 (from January 1, 2027), the EU AI Act, NYC Local Law 144 and the biometric consent laws, so the work compounds rather than duplicates.

Done well, Portland readiness is a small, concrete step within a broader AI compliance program, and a good AI compliance tool makes it repeatable across every location you operate. To see which of these laws apply to your own footprint and what that would cost, visit https://aicompliant.ai/pricing.

Frequently Asked Questions

What does Portland's private face recognition ban (City Code ch. 34.10) prohibit?

It prohibits using face recognition technologies in places open to the public, such as stores, restaurants, hotels and venues. The ordinance was enacted as Ordinance 190114 and has been effective since January 1, 2021. Confirm the current text and any exceptions with Oregon counsel.

What are the penalties for violating the Portland ban?

Enforcement is through a private action under City Code §34.10.050. A plaintiff can recover damages sustained or $1,000 per day for each day of violation, whichever is greater, plus other appropriate remedies. A prevailing plaintiff who made a written demand at least 30 days before suing can also recover attorney fees.

Not on the basis of the requirement summarized in our catalog. Illinois BIPA requires a written policy, notice and written release before collecting biometrics, and Texas and Washington require notice and consent. Portland's rule is a prohibition on use in places open to the public, so the compliance control is turning the capability off, not collecting consent.

Does Portland's ordinance affect our Colorado AI Act or EU AI Act compliance work?

They are separate laws, but the preparation overlaps. Colorado SB 26-189 (formerly SB 24-205) applies from January 1, 2027, and the EU AI Act's high-risk obligations apply from 2 December 2027 (Annex III) and 2 August 2028 (Annex I). The AI system inventory you build for Portland is the same foundation those programs need.

How can software help with a face recognition ban?

An AI compliance platform can maintain a location and system inventory, track vendor responses, and draft compliance documents for counsel's review. It cannot decide whether a particular deployment is lawful, so legal review remains essential.

Check if this regulation applies to your business

Use our free compliance checker to see which AI regulations apply to your company based on location, industry, and AI systems.

Free compliance checker →

← Back to blog