San Francisco became a lab for police surveillance after initially fighting it; the SFPD had a record 700 drone flights in February, up from 93 in February 2025
In 2019, the Board of Supervisors passed a landmark law making San Francisco the first major U.S. city to block the use of facial recognition technology by law enforcement.
Context & Ripple Effects
San Francisco’s current drone expansion reverses the city’s earlier posture as a privacy-policy pioneer: the Board of Supervisors backed a 2019 ban on law-enforcement facial recognition, then later amended the surveillance framework to permit Face ID on city-issued phones.
The jump from 93 SFPD drone flights in February 2025 to 700 this February makes the city a consequential test case for whether technology-specific restrictions can constrain a broader expansion of police surveillance tools. California’s earlier police-drone experience has already documented resident privacy concerns around sustained aerial operations in Chula Vista’s drone program.
First-order effects
- SFPD’s much higher flight volume expands the department’s immediate capacity to collect aerial footage during police operations.
- San Francisco residents face substantially more frequent police-drone activity, intensifying the practical importance of operating rules, public disclosure, and oversight.
Second-order effects
- The increase puts pressure on the city’s surveillance-governance framework to address drones as a distinct collection channel rather than relying on a facial-recognition ban alone.
- It gives other police departments and drone vendors a prominent municipal example of how quickly routine drone use can scale, while civil-liberties advocates gain a clearer volume metric on which to challenge safeguards.
Third-order effects
- If this pattern persists, local surveillance policy will increasingly turn from yes-or-no bans on individual technologies toward governance of how systems are deployed, logged, retained, and reviewed in combination.
- The central policy divide may shift from facial recognition specifically to the cumulative monitoring reach created when aerial collection becomes routine, even where earlier privacy restrictions remain formally intact.
The trend: Police surveillance governance is moving from narrow restrictions on named tools toward scrutiny of the operational scale and combined reach of expanding sensor networks.