AI-Written Software Turns a Cheap Drone Into a Facial-Recognition Stalker

3 articles · Updated · NBC News · Aug 5

A cheap consumer drone tracked a person around a San Francisco house using facial recognition, flying largely without direct human control and slipping through doorways while avoiding lamps.
Widely available AI models from Anthropic, OpenAI and others were used to build the software directing the drone, showing how off-the-shelf tools can automate target pursuit.
The demonstration centers on consumer-grade hardware rather than specialized military systems, underscoring how advanced autonomous surveillance can be assembled from accessible technology.

Sources

Left100%

NBC News78d ago

AI-Powered Drone Stalks Human Target Using Facial Recognition in San Francisco

ACLU78d ago

Fight Creepy ALPR Cameras | American Civil Liberties Union

Facebook78d ago

The experience of being chased by a murderous kamikaze drone, a flying machine that behaves as if it were alive, is not one that humans are equipped to understand. In the Atlantic:

If frontier AI turns toy drones into autonomous hunters, what stops this off-the-shelf technology from becoming the ultimate untraceable surveillance tool?

How close are we to a reality where cheap, AI-powered drones can autonomously stalk anyone, anywhere, without human oversight?

AI Drones Go Domestic: The 2024–2026 Global Surge in Autonomous Surveillance and the Race to Regulate

Overview

This report traces how AI-powered drones, first developed for military use in conflicts like Ukraine to overcome electronic jamming, have rapidly shifted into domestic surveillance tools for police and government agencies. In cities such as London and Las Vegas, drones equipped with live facial recognition and advanced tracking software are now used to locate suspects and monitor public spaces, often exploiting legal loopholes and private funding. The spread of cheap hardware and open-source AI has also enabled private individuals and corporations to deploy their own surveillance drones, blurring the line between public and private monitoring. However, these systems face technical vulnerabilities, such as GPS spoofing and algorithmic bias, leading to wrongful arrests and privacy violations. As legal frameworks struggle to keep pace, grassroots movements are emerging to demand transparency and accountability, highlighting the urgent need for robust regulation and oversight.

...