AI Being Used for Fake 911 Calls
High tech sadly comes with high tech scams
11/16/20262 min read
At the beginning of the school year, a caller made a bomb threat to Wheatland Union High School in Yuba County, Calif. Police responded overnight and the decision to close classes was made. The reality was… there was no bomb, and no one was hurt. It was part of a series of hoax calls - sometimes called swatting - to schools across the country.
Such calls can be made to either 911 or police departments. Callers often report a bomb threat, a mass shooting, or some other life-threatening emergency, all with the intention of creating mass hysteria and huge emergency response.
Prank calls have long been a problem, but the incident in Yuba County was somewhat new. This particular hoax call was made by an AI-generated voice. It’s becoming a new and growing challenge for emergency responders. False calls usually trigger a big, chaotic response from police, redirecting safety resources from real emergencies. In some cases, people also have been killed or injured in swatting attacks.
According to Mel Maier, CEO and executive director of Association of Public Safety Communication Officials, "Those [fake] voices used to be pretty easy to recognize, but now they're natural sounding. They can insert pauses. They can use words like, 'uh', or 'maybe, um, let me think about it.' And it sounds very natural."
Public Safety Threat Alliance, a cyber-security group that shares its findings, looked into the prevalence of swatting calls to 911. Based on 766 responses from public safety communications professionals, about 14% of dispatchers said they were targeted using AI tools. Many dispatchers believe that number is going to keep going up with increasingly advanced tools.
ElevenLabs, an artificial intelligence company which creates synthetic voice software, offers AI tools that are low-cost and easily available. Using the company’s software, one can change how deep or high-pitch their voice is. Pauses can be made shorter or longer, stutters can be added, and it all can be done for less than $20.
The ever-increasing technological sophistication of fake voices, and even background noises, makes it extremely difficult for emergency dispatchers to deduce what is real, or not real, during "high-stakes, high-consequence" calls.
There are, however, some hallmarks of a swatting call. For example, most swatters do not dial 911 directly. They call police departments, or even courthouses, who then transfer the call to 911. The reason being, direct calls to 911 show what number the call is being made from. By calling elsewhere first, it takes time to find the originating number. That’s time not available for a true emergency call situation to be evaluated.
New methods of training dispatchers are being created, most of which are not being made public. However, one method is to ask questions whose answers are unlikely to be found in a public database, and hopefully, not likely to be on a social media post. I have my doubts as to that approach. Almost everything can be found in databases and on social media.
Rest assured, dispatchers will always assume emergency calls are real and send out emergency resources.
Source used: NPR
