Pflugerville adopts SAFE policy to tier AI and surveillance risks

A new framework for city tech

On October 27, 2026, the Pflugerville City Council reached a unanimous agreement to implement the Secure, Accountable, Fair and Explainable (SAFE) Technology policy. This new set of rules dictates how the Texas municipality manages artificial intelligence, surveillance equipment, and biometric identification. It also sets the standard for how the city handles and shares data.

Unlike some local ordinances that ban specific products, the SAFE policy does not bless or condemn any single piece of software. Instead, it sorts all city technology into three distinct risk tiers based on what a tool is meant to do. This classification determines who has the final authority to sign off on the technology's use.

The policy was designed to balance the benefits of modern services with the need for public trust. Mayor Doug Weiss stated that technology should help the community without coming at the expense of privacy or accountability. This approach aims to provide a clear roadmap for how the city adopts new tools while protecting resident information.

How the risk tiers function

The spine of the SAFE policy is its risk tiering system, which separates routine operations from high-impact activities. Low-risk tools are considered routine administrative and operational items. These stay under the authority of the city manager, allowing staff to renew software without needing a public vote for every item.

The moderate tier exists as a middle ground, though specific details on which tools fall into this category are not yet fully defined. This tier is where tools that are not obviously surveillance but still touch resident data would likely reside. Records requesters may eventually probe this gap to see how the city classifies non-obvious data processing.

High-risk technology is the most strictly regulated category under the new rules. This tier includes systems used for law enforcement, surveillance, and biometric identification. Any tool falling into this category requires a formal review and approval by the City Council before the city is allowed to use it.

Human oversight and accountability

A core component of the SAFE policy is the requirement for human responsibility in automated processes. Council member Jonathan Coffman noted that the policy requires a person to remain responsible for decisions made using AI. This means a model's output can inform a city decision, but a human must ultimately own the outcome.

This rule is intended to prevent automation bias, where humans might blindly follow whatever a system suggests. By requiring a human to stand behind a decision, the city ensures that a named employee can be held accountable. This is particularly important for residents who may need to challenge a specific outcome or action taken by the city.

The 'Explainable' part of the SAFE acronym also points toward a need for transparency in procurement. If a vendor cannot describe how its model reached a specific output, the city department may struggle to meet the policy's goals. This puts pressure on software providers to offer clear insights into their underlying logic.

Transparency and future-proofing

The SAFE policy includes provisions for public transparency regarding the city's technology inventory. Residents will be able to see what technology the city uses and how their information is handled. While the specific format is not yet set, this creates a mechanism for residents to audit the city's digital footprint over time.

The policy is also designed to be forward-looking. It governs not just current tools, but any technology the city considers using in the years to come. This means a tool that does not exist yet still falls inside the framework once the city begins to evaluate it.

The policy uses the phrase "comparable high-impact uses" to define the top risk tier. This suggests the tier is defined by the effect on residents rather than a fixed list of products. A vendor should not assume a tool is exempt simply because it is not specifically named in the current text.

Guardrails for future tech

The reporting indicates that the policy may include a prohibition on facial recognition and biometrics, though this is currently described by council members rather than quoted from the text. Because biometric identification is confirmed as a high-risk category, the council holds the power to approve or deny specific uses. This creates a pre-use gate where the debate happens before deployment rather than after.

The policy was designed to balance the benefits of modern services with the need for public trust. Mayor Doug Weiss stated that technology should help the community without coming at the expense of privacy or accountability. This approach aims to provide a clear roadmap for how the city adopts new tools while protecting resident information.

The policy was designed to balance the benefits of modern services with the need for public trust. Mayor Doug Weiss stated that technology should help the community without coming at the expense of privacy or accountability. This approach aims to provide a clear roadmap for how the city adopts new tools while protecting resident information.