70% Error Rate: Are Police AI Cameras Flagging Wrong Drivers?

And welcome back to the factor on since it could high tech police cameras be making massive mistakes out there. A report reveals flock a one cameras, artificial intelligence cameras, misread license plates more than 70% of the time. In some cases, falsely flagging innocent drivers. This data comes right as a League City City Council unanimously voted tonight to let voters decide the future of the city's flock cameras this November. Joining us to talk about it. Senior Sergeant Chris men's men's air and Lieutenant Ashton Johnson Hall Hall and Harris of Harris County Precinct seven. Okay, I'm not even reading right tonight and Houston City Council member. I bet not miss him this name. Dr Carolyn Evans. Shabba. So when we talk about flock cameras or artificial intelligence cameras because flock is just a name brand. Where do you stand on that for the city of Houston? Actually, I think it's a time that we're gonna have to look to see what kind of errors and fix it because we certainly don't want information going out erroneously that are putting people in jeopardy. And especially in these times, you know, where people are getting stopped and deported and all of those things. We need that technology to work and it work properly. And if it doesn't, then we need to go back to the drawing board because I do believe that cameras are very beneficial, not just to catch criminals, but to deter crime. And so I don't want to throw the baby out with a bath water, so to speak. But I do think that they need to tighten up to make sure that this technology is working appropriately and being used appropriately. All right, Sergeant Manzanar your thoughts about cameras. Is there a place for them in society? Do you think they're useful? Are there too many mistakes? Yes, I do believe that they are useful, especially for us law enforcement to utilize as a tool. And that's all it is for us as a tool. Uh, not enough to stop a vehicle, but there are parameters in place like a visual verification of the return. Verify that the plate matches that vehicle. Do a secondary check through the National Crime Information Center that you're looking at the correct vehicle before you go and try to make contact with it. All right, Lieutenant Hall, there's been a case, a high profile case where a woman was identified as a suspect in a murder twice by these artificial intelligence cameras. Uh, is it an anomaly or still in your mind a good tool? So that case the flock has a has capabilities for us to enter a plate into the system and we can create a hot list for it. And those investigators added their vehicle to the hot list. So when the when the flash go off, the officers are are directed toward that area to stop the car. So it seems there's some human error there from the investigators that need the vehicle stopped anymore or they didn't have any more questions for the first person driving. They should have went in there and pulled that information or that hit from the system. The person who stopped and accused of being a suspect in a murder. Do they care if it's human error or A. I. Not at all. They want to make sure that you are not targeting them. They want to make sure that the information that you have gotten is appropriate. And so, you know, and then we have to realize that flock is not the only device that can be used. And so if they don't get themselves in order in regards to giving out appropriate data, then we need to look at some other sources because we never want people to be stopped, especially when they are innocent. And we know how things can escalate even if you're an innocent person. So we're trying to cut down as much police interaction erroneously as possible. So there is a responsibility before using this technology to make sure that it is operating appropriately and being used appropriately. And I do realize that they're going to be voting in Lake City. I never have a problem with the citizens making a decision, but make sure that they are looking at a right decision in terms of whether these cameras are operating for the purpose for which they are intended and on the lines of that used for the purpose that they're intended. We've had this for any of you. We've had at least 50 officers around the country who have been charged as a result of misusing these cameras in instances of personal family relationship issues. Is that a concern for any of you? Let's hear from you can't legislate integrity. You don't have integrity. I don't care what you have and what tools you have at your disposal. And so we have to do good background screening to make sure that people are in the jobs that they're supposed to be in and that they're going to operate in the manner in which they were hired. And so, you know, it's very hard to regulate that because sometimes people get in and then they become corrupted because of the access. And so we have to be mindful of that and always keep an eye to make sure that people are doing the job in the manner that they are intended to do. All right, Lieutenant Sergeant, your thoughts on that. Is it temptation? Is the temptation too great for some officers when they see opportunity and there's a chance to check on your side piece or your wife or your husband? Unfortunately, I believe the flock system officers by any so much that we have these, uh, these these negative interactions with the system with those bad officers. Uh, not to say that because there are thousands of officers across America. We're talking about 50 now, right now, 50, right? What? Wait, no, it's never right. It's never okay to, uh, misuse any law enforcement system. Uh, you know, the flock system, like I said, it's so well that, you know, these officers feel that, okay, let me engage with my my work tools to look for my look at my side. She's on their wives or whatever, whatever it is they're looking for. All right, Sergeant from this. Oh, absolutely. And these officers that I'm referring to, they've all been charged for that. What can you tell the public to keep their confidence when it comes to these cameras, Sergeant? So, like I said earlier, there are steps to verify that return. We can't just look at flogging. So let's go stop the vehicle. We have to take steps, visualize the plate, make sure there's no discrepancies within the plate that we got compared to the actual picture. Look at the vehicle and check it through the N. C. I. C. System. Make sure it's the exact vehicle. And then you on top that you want to develop an additional reason to stop the vehicle, right? But not just solely based on the return. We don't really do that. We have steps to verify that that is the accurate vehicle. All right. We want to thank you all for joining us here on the factor on Simpson