Artificial intelligence (AI) isn’t a tool of the future—it’s a present-day need. In a time when technology is getting smarter by the day, it’s imperative that law enforcement agencies (LEAs) have the opportunity to access technology that can help them perform their duties more efficiently and more accurately.
With digital evidence involved in roughly 90% of criminal cases, including surveillance, video, audio and social media, LEAs need modern solutions to overcome the burden of manual review.
When using the proper tools, we see that AI becomes an invaluable productivity assistant. It can save staff time, optimize workflows, and allocate more resources toward human-centric tasks that cannot be automated.
Modern AI systems used in law enforcement are built on machine learning models that analyze large volumes of data inputs from multiple data sources. These AI technologies support intelligence gathering and investigative decision-making, but AI outputs must always be reviewed by human investigators as part of responsible human oversight within the criminal justice system.
As a part of Veritone’s commitment to helping law enforcement agencies improve workflows and increase transparency and trust with the communities they serve, we’ve developed a series on how AI can help relieve existing pain points and improve day-to-day functionalities.
How is law enforcement using AI?
It may sound like something reserved for big tech companies, but AI can be used in a wide range of functional tools that law enforcement agencies can leverage to improve efficiency and accuracy. These real-world applications provide clear examples of AI in law enforcement, from evidence analysis to suspect identification and digital redaction. While AI doesn’t (and most likely never will) have the power to replace human abilities, what it can do is automate tedious, time-consuming tasks and prepare data for human review, insight, and interaction.
In a nutshell: the right AI-infused tools can help law enforcement agencies maximize productivity and perform tasks at scale so they can dedicate more attention and resources to human-centric tasks.
Artificial Intelligence for Audio & Video Redaction
Veritone’s AI solutions for law enforcement were created to automate the back office tasks that take dozens of employees and hundreds of man-hours to complete so law enforcement’s human efforts can persist.
One tool that can accelerate investigations and help process evidence more efficiently is Veritone iDEMS, a comprehensive suite of AI-driven applications that create an intelligent digital evidence management system, enabling rapid access, efficient search, seamless organization, and secure sharing of digital evidence, all designed to streamline investigations and strengthen legal processes.
These investigative capabilities are supported by AI models trained on diverse crime data, allowing police departments to identify patterns across large volumes of unstructured information. Together, these deployments represent practical examples of AI in law enforcement that strengthen investigative workflows while keeping human intelligence at the center of evidentiary decisions.
Eighty percent of the world’s data is unstructured, (i.e. video, audio or image files, log files, social media posts, etc), so it’s no surprise that huge amounts of evidence are also unstructured. And since unstructured data in particular is so time-consuming to manually search through and analyze, it’s not uncommon that large amounts of digital evidence end up not being reviewed simply because the volume of work required to manually analyze it would drain department resources.
Sifting through evidence manually puts immense pressure on departments, but Veritone iDEMS culls down evidentiary media by centralizing it in a secure platform before enabling AI to help make it more searchable and discoverable. This means that investigators can review relevant data in a fraction of the time.
Veritone Illuminate can also:
- Generate actionable insights from unstructured evidence using advanced AI analytics.
- Transcribe and translate audio and video files to enable searchable keywords and phrases.
- Facilitate collaboration and evidence sharing across teams within a centralized, secure hub.
Beyond streamlining evidence analysis, Veritone’s iDEMS suite includes tools to further accelerate investigations, including Veritone Investigate, Redact, Illuminate, Track, and IDentify, all integrated in a single AI-powered digital evidence management system.
Artificial Intelligence for suspect identification
Access to countless hours of video evidence can help LEAs identify suspects and solve cases, but many agencies still sift through these seemingly endless loads of footage manually, which occupies a lot of department resources.
Not unlike the stored fingerprints of known-offender and person-of-interest records within LEA databases, booking photos of these individuals can also be used as a foundation to compare against collected video or still-image evidence, enabling more operational efficiency and accuracy among departments.
AI-powered facial recognition technology and DNA analysis are increasingly used to support suspect identification and missing persons cases. While these tools can improve speed and accuracy, their use within the criminal justice system raises ethical implications related to biased data, civil liberties, and the need for careful consideration and human oversight in how AI algorithms are applied.
By leveraging AI solutions such as Veritone IDentify to help them identify suspects quicker than ever before, police can streamline the process of suspect identification and allow the software to analyze previously booked and arrested individuals recorded in their own databases.
Veritone IDentify can also:
- Create virtual “six pack” line-ups of potential suspects for further evaluation
- Share suspect lists with intra- and inter-agency colleagues
- Organize evidence by case including case details such as ID, department, officer(s), description, etc.
Using IDentify optimizes investigative workflows, allocates human hours to more important tasks, and significantly reduces costs for law enforcement agencies. There are other time-saving opportunities for LEAs to use AI while processing ESI, particularly in regard to redacting sensitive information.
Artificial Intelligence for Audio & Video Redaction
There is a demand for both public justice and public safety agencies to provide audio, video, and image-based evidence with redacted sensitive information. Redaction ensures the safety of witnesses, complies with freedom of information laws, and meets requirements for public records requests. The process of manually redacting this information is both time-consuming and expensive for agencies—so how can applications of AI in law enforcement help?
Rather than dedicating dozens of employees and hundreds of man-hours to manual redaction, agencies can use AI-infused tools like Veritone Redact to detect and redact specific items (like heads or license plates) in a quick, cost-effective manner.
After the agency has redacted audio, video, or image-based evidence, Redact allows staff to download the files to local computers for further processing.
Veritine Redact can also:
- Manage digital evidence redaction workloads by tagging audio and video evidence with its status in the approval workflow
- Capture comprehensive reporting of all actions taken against redacted digital evidence to support compliance with chain of custody requirements
With AI-assisted redaction, identification, and discovery, public safety agencies can increase efficiency, improve investigative workflow, and optimize their day-to-day operations while increasing transparency and building better relationships with their community.
Should agencies and teams need further assistance with managing the redaction workload from a third party, Veritone offers Redaction Managed Service.
Artificial Intelligence for Tracking Persons of Interest
In a world where mountains of data can be created in a matter of seconds, video footage from multiple sources and angles is a double-edged sword for investigators. On one hand, the likelihood of finding footage of a suspect, witness, or victim increases drastically, but it could take several hours and employees to comb through and find them.
In addition to video analytics, some agencies integrate social media monitoring tools and other data sources to support intelligence gathering and threat assessment. AI-powered anomaly detection can alert police to unusual activity patterns across multiple data streams, helping teams focus investigative resources more efficiently when time is critical.
With Veritone Tracker, agencies and investigators have a greater ability to quickly track individuals across videos, which can help teams locate persons of interest (POI) across several videos without the use of personally identifiable information (PII), so teams will remain within compliance with privacy laws.
Key features of Veritone Tracker include:
- Human-like object (HLO) Identification builds the necessary criteria for AI to compare against other video files by identifying and defining human-like objects in a video frame.
- Confidence-based similarity detection doesn’t use facial or biometric recognition, which maintains compliance with privacy laws and protects the PII of those involved.
- Multi-source video analysis speeds up the video analysis process while collecting valuable insights.
In situations where time is of the essence, AI can help agencies find critical information so investigators can focus their energy on more crucial efforts.
Artificial Intelligence for evidence management
The rising volume of digital data has led to challenges for LEAs, investigators, and other teams handling digital evidence. These challenges range anywhere from workflow inefficiencies and backlogs to burnout and budget constraints.
Veritone Investigate addresses these issues by offering an AI-powered Digital Evidence Management System (DEMS) that enables public safety agencies and organizations to efficiently access, search, manage, organize, and share digital evidence.
Veritone Investigate facilitates faster investigations and strengthens legal processes by allowing users to:
- Seamlessly switch between apps, including Veritone Redact, iDentify, Tracker, and Illuminate
- Manage, organize, share, and distribute digital evidence
- Instantaneously enrich files with metadata, including unstructured data
- Automate and optimize workflow within a custom interface
This platform operates within one secured CJIS-Compliant cloud environment, and, unlike conventional systems, Veritone Investigate is powered by AI tools that work with a comprehensive suite of solutions—allowing LEAs to analyze, manage, prepare, and share digital evidence in one safe and secure central hub.
Artificial Intelligence for Predictive Policing
Predictive policing technologies use machine learning to analyze historical crime data and other datasets to identify patterns and forecast potential future crimes. These predictive policing systems help police departments allocate resources more proactively by identifying crime hotspots and areas of elevated risk, supporting more proactive policing strategies.
Some predictive policing tools integrate data from license plate readers, gunshot detection systems, and vehicle surveillance systems to improve situational awareness and alert police to emerging threats in near real time. For example, police departments such as the Los Angeles Police Department have explored data-driven approaches to crime prevention using AI-powered predictive policing tools.
However, predictive policing raises growing concern about biased data and potential risks to civil liberties if AI algorithms reinforce existing disparities in the criminal justice system. Ethical guidelines, impact assessments, public input from community members, and strong human oversight are essential to ensure predictive policing technologies are applied responsibly.
Building Trust and Transparency with the Community
Over the last few years, an important discussion about policing has arisen throughout the US. In our 2021 Transparency and Trust report, we surveyed 3,000 Americans and found encouraging responses:
- 4 out of 5 Americans feel safe in their communities and most credit the police
- 42% say lack of transparency has hurt their perception of law enforcement
- 84% say police should focus on responding to violent crime matters
- 61% say they trust the police to use technology to aid in suspect identification
To address concerns around AI technologies in policing, agencies must establish clear governance frameworks that include ethical guidelines, impact assessments, and transparent development processes for new systems. Engaging community members and incorporating public input into AI deployment decisions helps ensure these technologies support public safety while respecting civil liberties within the broader criminal justice system.
Public demand for transparency is higher than ever before, which has put a strain on law enforcement man-hours, but fortunately, technology is now mature enough to help LEAs create more transparency at a scalable level.
The relationship between law enforcement and technology can be strenuous: over the past few decades, there has been a boom in tech that could help with investigations and identifying suspects, like footage from CCTV or cellphones, but there weren’t many tools that could help departments keep up with the evergrowing needs for data management and processing.
With 90% of law enforcement professionals supporting the use of AI within their agencies, and another 87% believing that AI is transforming public safety for the better, AI-infused tools like Veritone Illuminate, IDentify, and Redact can help meet the challenges LEAs face today.
In this series:
How Law Enforcement Builds Transparency and Trust With Technology
This blog will dive into Veritone’s Trust and Transparency Report and discusses how technology can bridge the gap between LEAs and the communities they serve.
Racial and Identity Profiling in Law Enforcement Agencies
States with laws regarding racial and identity profiling in law enforcement may unintentionally create a heavier workload for officers. Learn about how AI technology can help optimize this workflow.
Processing FOIA Requests With AI
Discover how AI technology can help law enforcement agencies keep up with the demands of Freedom of Information Act requests in a timely manner.
How AI Can Help Law Enforcement Identify Suspects Faster
Learn how Veritone solutions for law enforcement can assist with faster, safer, and more accurate suspect identification.
