How gait tracking works
Instead of relying on a plate number, OS Investigate uses the way a person walks to pick them out. The tool analyzes movement patterns from camera footage, letting officers follow a suspect even if the vehicle has no visible plate or the driver isn't known. The underlying code includes 69 preloaded AI prompts, which shape how the system interprets what it sees.
Beyond license plates
Flock's cameras have long been used to read plates and flag stolen cars. OS Investigate pushes that further. With this tool, a person's body movement becomes the identifying marker. That means a driver can be tracked across multiple cameras without any other identifying information.
What the prompts reveal
The 69 prompts are baked into the software, but Flock hasn't published the full list. They're designed to guide the AI in recognizing specific movements and behaviors. The exact range of what the system can flag isn't public.
Unanswered questions
How police departments will use the gait data, and what safeguards exist, remains an open question. Flock hasn't said which agencies have adopted the tool or when it will be rolled out. The technology is new, and so are the privacy questions that come with it.
That's about 250 words. Need to expand to 500-800. Let's add more detail and context. We can talk about the implications but not attribute. We can say "The tool raises concerns about surveillance" but that's opinion. We can say "Civil liberties groups have not commented" - but we can't invent. So we can say "The tool is a departure from Flock's previous focus on license plates" - that's factual based on known info? We know Flock is known for plate cameras, so that's fine. Let's write more. Also need to ensure no fake quotes. We'll paraphrase. Let me write a full draft: Title: "Flock's New AI Tool Tracks Drivers by Gait, No License Plate Needed" Lead: Flock, the company known for its license-plate-reading cameras, has introduced a new AI tool for police that can track a driver by the way they move. Called OS Investigate, the system works even when there's no license plate or name to go on. It turns Flock's cameras into a surveillance network that identifies individuals by their gait — the unique pattern of their walk. Section: "A shift from plates to people" - talk about how Flock's cameras were originally for plates, now they can do more. Mention that the tool uses movement patterns. Section: "Inside the 69 prompts" - explain that the code includes 69 preloaded AI prompts. These prompts guide the AI in recognizing patterns. The company hasn't released the list, so it's unclear what specific behaviors are flagged. Section: "How police might use it" - we don't have specific uses, but we can say it could help find suspects without a plate. We can say "The tool could let officers follow a suspect across a city without knowing who they are." That's inference but reasonable. Section: "Questions about privacy" - we can say the technology raises questions, but we can't attribute to experts. We can say "The tool's reliance on gait data is a new frontier for surveillance" but that's editorial. We can say "It's unclear what happens to the gait data after it's collected." That's a question. End with: "Flock hasn't said when the tool will be available to police departments or which agencies have signed up. The company has not released the full list of 69 prompts. Those details will likely shape how the tool is used — and how much it's challenged." That's a concrete next step. Let me write it out. Also, ensure we use contractions: "it's", "doesn't", "won't". Vary sentence length. Let's write the final article in the content field. Word count target: 600-700. I'll write: Flock, the company known for its license-plate-reading cameras, has introduced a new AI tool for police that can track a driver by the way they move. Called OS Investigate, the system works even when there's no license plate or name to go on. It turns Flock's cameras into a surveillance network that identifies individuals by their gait — the unique pattern of their walk.A shift from plates to people
Flock's cameras have long been used to read license plates and flag stolen cars. OS Investigate pushes that further. Instead of relying on a plate number, the tool analyzes how a person walks. That means a driver can be followed across multiple cameras without any other identifying information. The underlying code includes 69 preloaded AI prompts, which shape how the system interprets what it sees.
Inside the 69 prompts
The 69 prompts are baked into the software, but Flock hasn't published the full list. They're designed to guide the AI in recognizing specific movements and behaviors. The exact range of what the system can flag isn't public. That leaves a lot of unknowns about what the tool is capable of.
How police might use it
The tool could let officers track a suspect without knowing their name or plate number. It's a capability that goes beyond what Flock's existing products offer. But it also raises questions about how far surveillance can go. A person's walk becomes a digital fingerprint, and that fingerprint can be followed.
Questions about privacy
It's unclear what happens to the gait data after it's collected. Whether it's stored, shared, or deleted isn't known. The company hasn't said which police departments have adopted the tool or when it will be rolled out. Those details will likely shape how the tool is used — and how much it's challenged.
That's about 300 words. Need more. Let's expand with more specifics. We can talk about the fact that it doesn't need a name or plate, so it's useful for hit-and-runs or suspects who cover plates. We can say "The tool works by analyzing movement patterns from camera footage" - we already have that. Maybe add a paragraph about the technical side: "The prompts are preloaded into the code, meaning the AI starts with a set of instructions. How those prompts are applied in real time is not disclosed." That's fine. We can also add a sentence about the company's background: "Flock is best known for its network of cameras installed in neighborhoods and along roads." But that's not in facts, we can say "Flock, the company known for its license-plate-reading cameras" - that's already in lead. Let's write a longer version. Also, ensure we don't use "It's worth noting" etc. Let me craft final content:Flock, the company known for its license-plate-reading cameras, has introduced a new AI tool for police that can track a driver by the way they move. Called OS Investigate, the system works even when there's no license plate or name to go on. It turns Flock's cameras into a surveillance network that identifies individuals by their gait — the unique pattern of their walk.
A shift from plates to people
Flock's cameras have long been used to read license plates and flag stolen cars. OS Investigate pushes that further. Instead of relying on a plate number, the tool analyzes how a person walks. That means a driver can be followed across multiple cameras without any other identifying information. The underlying code includes 69 preloaded AI prompts, which shape how the system interprets what it sees.
The tool works by analyzing movement patterns from camera footage. It doesn't need a name, a plate, or even a clear view of a face. That makes it useful for cases where a suspect has covered their plates or is driving a vehicle that isn't registered. But it also means the system can pick out anyone who crosses a camera's path, not just someone already under investigation.
Inside the 69 prompts
The 69 prompts are baked into the software, but Flock hasn't published the full list. They're designed to guide the AI in recognizing specific movements and behaviors. The exact range of what the system can flag isn't public. That leaves a lot of unknowns about what the tool is capable of. Some prompts might be aimed at identifying a running person, a limping individual, or someone carrying something heavy. Without the list, it's hard to know.
How police might use it
The tool could let officers track a suspect without knowing their name or plate number. It's a capability that goes beyond what Flock's existing products offer. But it also raises questions about how far surveillance can go. A person's walk becomes a digital fingerprint, and that fingerprint can be followed. The company hasn't said which police departments have adopted the tool or when it will be rolled out.
Unanswered questions
It's unclear what happens to the gait data after it's collected. Whether it's stored, shared, or deleted isn't known. Flock hasn't released the full list of 69 prompts, and it hasn't detailed the safeguards around the tool. Those details will likely shape how the tool is used — and how much it's challenged.
That's about 400 words. Need more. Let's add a paragraph about the company's history or the significance. We can say "This is the latest step in Flock's expansion from simple plate readers to broader surveillance." That's inference but okay. We can also add a sentence about the technology: "Gait recognition is not new, but using it in a networked camera system like Flock's is." But that's an invented comparison. We can't say that because we don't have facts. Better to stick to the facts. Let's expand the existing paragraphs with more detail. For example, in the second paragraph, we can say "The tool is designed to help police find suspects faster, but it also means that anyone who walks past a Flock camera could be logged and identified by their movement." That's a logical extension, but is it factual? We don't know if it logs everyone. The tool is for tracking drivers, so it's likely targeted. We can say "The tool can identify individuals by their gait" - that's fact. So we can say "That means the system can pick out a specific person from a crowd based on how they move." That's reasonable. Let's also mention that it works without a name or plate, so it's a new capability. We need to reach 500-800. Let's write more. I'll write a final version: Lead: Flock, the company known for its license-plate-reading cameras, has introduced a new AI tool for police that can track a driver by
