By Jon Gacek, SVP and GM, Government, Legal & Compliance, Veritone, Inc., finalist of the Best SaaS Product For Law And Legal Services category at The SaaS Awards 2023 and the Best Digital Forensics Solution category at The Cloud Security Awards 2023

In the rapidly evolving domain of video evidence analysis, we find ourselves in the midst of a profound transformation, a paradigm shift that is revolutionizing the way investigative teams operate. These teams – ranging from justice and public safety agencies to private organizations – are constantly grappling with a record amount of video evidence, a volume so staggering that it transcends the capacity of traditional manual methods of review. In today’s world, countless hours of footage are generated, a trend showing no signs of abating.

Amid this video evidence explosion, it’s become necessary for legal organizations and public safety and justice agencies to harness the power of artificial intelligence (AI).

However, this technology doesn’t come without concern. As we continue to leverage the sophisticated capabilities of AI in investigations involving video files, there’s also an urgent, ethical challenge: how do we ensure the protection and preservation of personal identifiable information (PII)?

With the adoption of AI on the rise, privacy concerns are also increasing, bringing the conversation around how we can safeguard PII to the forefront. It’s a critical balancing act where we must weigh the demands of effective, comprehensive investigations against the imperative of privacy. Our responsibility, therefore, extends beyond the mere application of technology; it encompasses fostering an ecosystem where innovation and privacy are complementary rather than contradictory.

Overcoming the Explosion of Digital Media Evidence

“Best in Mobile” Cloud Solution category at The Cloud Awards

When you consider that an astounding 97% of Americans own some form of mobile device, this universal use of technology has drastically changed how evidence is gathered and analyzed, particularly in both the public and private sectors. Mobile devices have become a treasure trove of potential evidence, from smartphone videos taken by bystanders to footage captured on private security systems. This, in turn, has created a sea of video evidence, a dynamic, ever-changing landscape that promises to provide key insights while simultaneously being overwhelming to navigate.

In addition to the deluge of video footage sourced from personal devices, other compounding sources of video evidence have emerged, each one contributing to the escalating volume of data confronting investigative teams:

  • Security camera footage from many locations—from residential properties to commercial enterprises—offers its own wealth of evidence
  • Traffic camera footage provides a critical perspective on incidents occurring on our roads and intersections
  • Body-worn cameras, utilized by law enforcement and private security firms, offer a real-time, ground-level view of events as they unfold.

The compendium of these various sources results in a colossal and diverse collection of video evidence.

The Challenge of Multi-source Video Evidence

Predictably, this volume of video evidence is not static; it is destined to grow exponentially. As technology continues to permeate every corner of our lives, the prevalence of video recording devices—from traffic cameras to body-worn devices—will inevitably rise.

For instance, in the realm of law enforcement, the deployment of body-worn cameras has grown increasingly popular. Today, a significant 60% of local police departments and nearly 50% of sheriff’s offices have fully deployed the use of body-worn cameras. These devices offer invaluable real-time insights into the interactions between law enforcement officers and the public, acting as objective observers that record every moment.

As the adoption of these cameras becomes more pervasive, steadily inching closer to 100%, the volume of video evidence garnered from these devices is set to multiply, contributing significantly to the overall pool of video evidence that investigators have at their disposal.

Consequently, this creates a cascading effect: the more these recording devices are used, the greater the volume of video evidence generated.

Best Digital Forensics Solution category at The Cloud Security Awards

But this goes beyond public safety and justice. Video surveillance plays an integral role in a variety of situations and contexts. In sporting arenas, for example, video footage ensures spectator safety. Cameras deployed in these settings can capture crucial moments, help identify potential issues, and even prevent undesirable incidents. Similarly, within the vast networks of public transit stations, video footage can prove invaluable in identifying missing or kidnapped individuals. Every moment captured by these cameras could potentially hold the key to unraveling mysteries, solving crimes, and even saving lives.

Harnessing the potential of this torrent of video evidence, and navigating it effectively, is a formidable task. For investigators and analysts, the challenge lies not just in managing the sheer volume of data but in extracting meaningful insights from a sea of digital noise—a task that necessitates the transformative power of artificial intelligence.

Leveraging AI to Overcome the Burden of Evidence Review

80%, if not more, of crimes involve some form of video evidence. Sifting through this vast amount of video evidence to identify crucial details and connections is akin to finding the proverbial needle in a haystack. But, at the same time, it underscores the importance of effective and efficient video evidence review, bringing into focus the crucial role of technological innovations, such as artificial intelligence, in shaping the future of investigations.

Faced with the daunting task of meticulously reviewing massive volumes of video evidence, investigative teams are increasingly turning to technological solutions. At the forefront of these solutions stands AI. From simplifying the process of redacting sensitive portions in footage to aiding in the investigative review of multi-source evidence, AI solutions have begun to reshape how teams process evidence.

Consider, for example, the process of redacting a video. Traditionally, this could mean spending eight painstaking hours going through a 10-minute video frame by frame. But with AI, the same task could be accomplished in just a few minutes. In a similar vein, when tracking a person of interest across multiple video files and formats, AI can play a crucial role in accelerating and enhancing the process.

A human operator could identify objects in a video that are then searched for by the AI across the entire collection of video evidence, potentially surfacing moments where the identified individual appears. But this, too, presents potential privacy challenges. If AI is to be employed in this way, it must do so in a manner that respects the privacy of individuals, ensuring that personally identifiable information is not misused or exposed.

New AI Technologies to Offset Privacy Concerns

That’s the approach Veritone has taken in developing new AI solutions to tackle some of the toughest evidence-based challenges. Take facial recognition, for example. Technologically, it does not do well in processing large crowds. But ethically, it also raises concerns about privacy. But new AI technologies maintain a human on the loop to define human-like objects to surface moments where the suspect might appear across video files.

The profound potential of this technology lies in its capacity to enable rapid video evidence review while adhering to privacy laws and safeguarding PII. This, in turn, bolsters the role of AI as a key tool in expediting various processes, from swift criminal apprehension to efficient search for missing individuals, to effective countermeasures against human trafficking, and to building accurate timelines for cases.

By utilizing AI technology in this manner, we’re not just streamlining processes but setting the stage for a safer society. The time and resources saved can be reinvested in diverse areas to enhance how agencies serve their communities, fostering a future that respects individual privacy and the importance of swift, effective investigations. We are standing at the precipice of a new era where AI technology serves as a vital tool in our investigations and as a guardian of our privacy.