Pflugerville adopts SAFE tech policy to regulate AI and surveillance

A proactive approach to tech

The Pflugerville City Council voted to adopt the Secure, Accountable, Fair and Explainable, or SAFE, Technology policy during its meeting on October 27. This new framework establishes immediate safeguards and oversight for how the city handles artificial intelligence, surveillance tools, and data sharing. It serves as a proactive standard to guide the city's technological growth rather than reacting to problems as they arise.

The policy works by categorizing all city technology into three levels: low, moderate, or high risk. This classification is determined by the intended use of the tool and its potential impact on the community. By sorting tools this way, the city can apply different levels of scrutiny depending on how much power a specific technology holds over the lives of residents.

Council members noted that the goal is to create a living document that evolves with new developments. Mayor Doug Weiss stated that while technology helps the city serve the community, it should not happen at the expense of public trust. The policy aims to keep the people the city serves at the center of every technological decision made by local leaders.

Governance and oversight

One of the most significant components of the SAFE policy is the creation of a resident oversight commission. This group, appointed by the City Council, will review proposed technologies and provide formal recommendations to staff and elected officials. This ensures that the people living in Pflugerville have a direct voice in the tools being deployed in their neighborhoods.

The policy also distinguishes between routine administrative tools and high-risk systems. While the city manager will continue to oversee standard operational tools, high-risk technologies require direct City Council review and approval. These high-risk categories specifically include tools used for law enforcement, surveillance, and biometric identification systems.

Council member Rudy Metayer noted that the council learned many lessons from previous experiences with surveillance data. He stated that the council intends to ensure they have more governance and eyes on these things moving forward. This shift reflects a desire for more transparency and accountability regarding how public data is accessed and utilized by the government.

Restrictions on AI and data

The SAFE policy places strict limits on how AI can be used in government operations. While automated systems may assist city employees in their daily tasks, they are prohibited from replacing human judgment in significant decisions. This applies to critical areas such as law enforcement, code enforcement, the issuance of permits, and access to various city services.

Data privacy is another core pillar of the new regulations. The policy explicitly forbids vendors and outside parties from selling city data or using it for unrelated purposes. Specifically, third parties are not allowed to use city data to train general-purpose AI models unless the City Council gives express approval and the action is allowed by law.

Furthermore, the policy requires written agreements and City Council approval for any outside organization seeking ongoing or direct access to city data. There are limited exceptions built into the rules, but these are only permitted for specific emergencies or active public safety needs. This prevents the unauthorized use of information by external entities without a clear legal framework in place.

Surveillance and transparency

The policy includes specific prohibitions on several types of surveillance. Without authorization from the City Council, the city cannot use facial recognition or certain biometric identification systems. It also bans the generalized tracking of lawful First Amendment activity and social scoring systems that rank people based on their behavior.

To keep the public informed, the city will maintain a dedicated website and release an annual report. This report will identify high-risk technologies, describe how they are being used, and outline the safeguards currently in place. These disclosures will be subject to necessary legal and security limitations to protect sensitive information.

Council member Jonathan Coffman noted that the council will need to continue getting comfortable with these types of changing policies on a regular basis. He described the new rules as a great start for establishing a level of transparency for the public. This ensures that the community remains aware of the technological landscape as it evolves.

Lessons from the past

The adoption of the SAFE policy follows a specific incident involving a contract with Flock Safety. The city decided to end that agreement after a resident's review of the company's records raised concerns about who had the authority to search the data collected by the cameras. This experience highlighted the need for clearer rules regarding data access and surveillance.

Mayor Doug Weiss emphasized that residents deserve to know how these tools are used and what is being done to protect their information. He argued that the city must balance the benefits of new technology with the need for accountability and public trust. The policy is intended to bridge this gap by providing a clear roadmap for responsible innovation.

By establishing these rules now, Pflugerville aims to avoid the pitfalls of reactive governance. The policy ensures that any new technology must meet a standard of fairness and explainability before it is integrated into city operations. This move establishes a foundation for a more secure and transparent relationship between the local government and its residents.