Border Lockdown: Quantum Sensors & AI 🚨🛡️

Tech

🎧

Quantum Sensors on the Horizon
CBP is contracting with General Dynamics to develop a prototype of “quantum sensors” alongside a “database incorporating artificial intelligence,” intended to detect illicit objects and substances – including fentanyl – within vehicles, containers, and other devices. According to a contract justification published in the Federal Register last week, this project will integrate advanced quantum and classical sensing technologies with Artificial Intelligence, ultimately deploying proven concepts and end products across the CBP environment. The justification document states that CBP will take additional steps to enhance its ability to detect and, consequently, significantly reduce the harms associated with illicit contraband entering the United States of America, bolstering national security. The contract justification, which includes details of a $2.4 million General Dynamics contract made public in December 2025, obscures the name of the company developing the prototype.

AI-Powered Detection Strategies
CBP’s efforts align with a broader, department-wide push within the Department of Homeland Security (DHS) “to support the adoption and scaling of AI technologies,” as outlined in a strategy memorandum published last year. CBP and General Dynamics have not responded to requests for comment from WIRED. Last week’s justification for the project lacks specific details regarding the methods employed by its “quantum sensors” and the information processed by the AI database.

Leveraging Existing Detection Technologies
However, it does offer insights into the detection methods the agency has considered. The document outlines CBP’s market research, conducted between April and October of 2025. In July, CBP published an information request seeking a vendor to supply 35 handheld “Gemini” analyzers, sold by Thermo Fisher Scientific, designed to identify unknown chemicals and narcotics. DHS has previously tested the Gemini devices, according to reports published in 2021 and 2023. The July request, which specifies the substances the devices would analyze – including fentanyl, ketamine, cocaine, methamphetamine, diazepam, and MDMA – makes no mention of artificial intelligence or a database. The request states: “The detection equipment will be used by CBP Officers in non-intrusive testing to detect a wide range of narcotics, controlled substances, unknown substances, and general organic materials,” and notes that the agency “continues to seize an increasing number of opioids at the nation’s borders.” The document acknowledges that the devices could be prone to “false-positive and false-negative results.”

Exploring Quantum Chemistry for Enhanced Detection
Despite the ambiguity surrounding last week’s justification’s reference to “quantum” sensors, potential detection methods based in quantum chemistry exist, such as the use of “quantum dots” and fluorescent dye to detect fentanyl, as explained in a 2024 paper. Matthew Webber, an engineering professor and molecular science researcher at the University of Notre Dame and one of the coauthors on the 2024 paper, explained that the research utilizes “quantum dots”—artificially made, graphene-based nanomaterials. When combined with a fluorescent dye and a synthetic molecule functioning as a “basket,” these dots produce visible and quantifiable fluorescence. Specifically, Webber detailed that the addition of just a few micrograms of fentanyl causes the drug to adhere to the quantum dots, effectively “outcompeting” the fluorescent dye and resulting in a loss of fluorescence. He stressed that all fentanyl-related lab research is conducted within a rigorously controlled environment, never involving large quantities of the drug for legal, practical, and safety reasons. Webber emphasized the sensitive nature of this research and clarified that labs like his do not maintain bags of fentanyl, an impression that may be inaccurate.

AI’s Potential in Spectral Analysis
Furthermore, he suggested that artificial intelligence could potentially assist with tasks such as “spectral deconvolution.” “When dealing with signals coming from multiple agents within a mixture, the human eye may not be able to effectively deconvolve the spectra into individual components,” Webber stated. “However, AI-based specialized convolution frameworks hold significant potential in this context.”

This article is AI-synthesized from public sources and may not reflect original reporting.