The AI Insights Report on Retail Crime Prevention is an executive overview of the current state of artificial intelligence in Asset Protection, Loss Prevention, and Retail Operations. Analyzing qualitative data from across AP/LP and retail operations, the report is delivered as an un-gated interactive website and a downloadable PDF. It's updated quarterly, and the survey is open year-round. These insights are gathered through survey data, executive interviews, and original research conducted. No private, proprietary, or personal information is viewed or sourced as part of this report. Produced and managed by Telakus Inc in partnership with The D&D Daily.
Three years ago, OpenAI launched the first version of the ChatGPT mobile app. Consumer AI, what we're experiencing and watching in society now, is basically three years old. The race has been on since then, and it's accelerating into your world of retail crime prevention. At the core of this report is a transparent acknowledgment and belief. That is, AI is advancing too fast to deliver an "authoritative" insights report. That's why we've made it a dynamic experience, updated quarterly with the survey always on. It's deliberately void of academic research standards and quantitative data. Instead, we're pairing industry sentiment with manual research based on publicly available information. And importantly, it doesn't require a form submission to view.
“The AI economy is changing every profession. Some of the most innovative organizations and early adopters are in retail. As AI becomes more integrated with retail crime prevention, few areas of the enterprise will be more complex than loss prevention and asset protection — and I believe few areas have more opportunity to grow and thrive than us too.”
Insights summary
Like many industries and segments within retail, change management is where the biggest focus is. In this context, most leaders are talking about (or not) having things like: AI Governance/Risk/Compliance (GRC) Policy, AI Enablement Program, Role-Specific AI Training, Location-Specific AI Training, and Product-Specific Training.
Video detection technology seems to be maturing the fastest in terms of adoption and willingness to buy. "Computer vision," while broad depending on your perspective, is the easiest category for leaders to reference and talk about. It's showing up at self-checkout, entries, and back-of-house. General reasoning is where the largest efficiency gains are reported. Leaders mention case narratives, policy checks, and investigator copilots.
Who we talked to & surveyed
- VP+24%
- Director/Senior Director41%
- Manager/Regional Manager24%
- Specialist/Associate11%
- Other/not listed0%
Industry-specific AI
“Organized Retail Crime groups are using technology to offend. It makes sense that loss prevention leaders are more connected with emerging technologies than ever before. Our community is proactive and innovative when it comes to outpacing the impacts of ORC, from detection and prevention to prosecution and closure.”
Alaina Kring · Senior Director, Partnership Development at ALTO- Video AI41%
- License plate recognition24%
- Person image matching18%
- Other AI tooling12%
- None deployed yet18%
Worth a deeper dive
If we drilled down on these categories, creating many more options to choose from, what are these 82% of leaders deploying that's not on this list (or is defined differently within this list)? It's important to note that these respondents are likely engaged with AI by self-selection. They chose to click on a survey about AI on RetailCrime.ai. The survey was also meant to be quick and easy, so it lacked definitions and context. This could impact response data due to differences in terminology and how technologies are packaged when deployed.
Ecosystem map
See and deter what happens in and around the store
- Alpha VisionActive
- PanoptycActive
- SAFRActive
- 3Si Security Systems
- Axis Communications
- Axon
- Cisco Systems
- CONTROLTEK
- Duress
- Everseen
- Facewatch
- Gatekeeper Systems
- Genetec
- HALOS
- Hanwha Vision
- i3 International Inc.
- Indyme
- Interface Systems
- InVue
- JVCKENWOOD USA
- LVT (LiveView Technologies)
- Motorola Solutions, Inc.
- OpenEye
- Real Time Networks
- RetailNext
- Reveal Media
- Sensormatic Solutions
- Skydio
- SmartSense by Digi
- Solink
- Spacture AI
- Spot AI
- Verkada
- Zebra Technologies
Frontier generative AI
- Daily GenAI use59%
- Weekly29%
- Occasional or never12%
- Top use: reports, emails, comms47%
- Top use: data analysis and search29%
Sentiment
“There's rightfully a lot of cautious optimism in our industry when it comes to AI. The truth about what's real today and what's possible tomorrow is something leaders are trying to get clarity on. Ultimately, I believe the right AI technologies and behaviors are a force multiplier in preventing retail crime.”
Hedgie Bartol, LPQ, LPC · VP of Retail Business Development at Alpha Vision- Accuracy/reliability of outputs29%
- Over-reliance on AI vs. human judgment29%
- Data privacy/confidentiality24%
- Legal/compliance risk12%
- No concerns6%
Glossary & Methods
AI Enablement+
AI enablement is the work of helping people and organizations use AI well, covering strategy, training, vendor selection, governance, and change management so AI tools produce real results in day-to-day work.
Computer Vision+
Computer vision is a field of AI that lets computers read and understand images and video, spotting objects, people, behaviors, and events. It powers technologies like video AI, license plate recognition, and image matching.
Facial Recognition+
Facial recognition in retail is the use of artificial intelligence and biometric cameras to capture facial geometry, convert it into a numerical dataset, and compare it against a database of known individuals for security or customer service. Facial recognition is a general technology that maps and compares facial features from digital images or video against a database, whereas live facial recognition (LFR) is a real-time application that scans active video feeds of crowds against a predetermined list to generate instant alerts.
Frontier Generative AI+
Frontier generative AI refers to the most advanced general-purpose AI models available, such as ChatGPT, Claude, and Gemini, which create text, images, code, and analysis from plain-language prompts and are not built for any one industry.
General Reasoning+
General reasoning is an AI model's ability to understand context, follow instructions, and work through problems across many topics. In AP/LP, it supports tasks like drafting case narratives, checking policies, and assisting investigators.
Governance/Risk/Compliance (GRC)+
Governance, risk, and compliance (GRC) is the set of policies, controls, and oversight an organization uses to make sure its use of AI is accurate, accountable, fair, secure, private, and in line with laws and internal standards.
Industry-Specific AI+
Industry-specific AI refers to AI tools built or trained for one industry's own problems, data, and workflows. In retail crime prevention, this includes video analytics, case management, and crime intelligence platforms made for asset protection and loss prevention teams.
License Plate Recognition+
License Plate Recognition (LPR) is a security and data-analytics system that uses high-resolution cameras and optical character recognition (OCR) software to automatically capture, read, and convert vehicle license plate images into searchable alphanumeric text.
Location-Specific AI Training+
Location-specific AI training is AI education that fits a particular place, covering local laws, privacy rules, store formats, and operations, such as one state, country, or store type.
Person Image Matching+
Image matching is a computer vision process that identifies, aligns, and compares similar visual elements, features, or entire regions between two or more images.
Product-Specific Training+
Product-specific training is instruction on how to set up, use, and get value from one particular vendor's AI product or platform, including its features, limits, and best practices.
RetailCrime.ai+
RetailCrime.ai is a free education and entertainment hub exploring how artificial intelligence is impacting asset protection, loss prevention, and store operations in their retail crime prevention efforts.
Role-Specific AI Training+
Role-specific AI training is AI education built around the duties, decisions, and risks of a particular job, such as an AP/LP investigator, a store manager, or a law enforcement partner.
Telakus+
Telakus is the publisher of RetailCrime.ai and a technology enablement platform for retail asset protection.
The D&D Daily+
The D&D Daily is the #1 daily digital news source for the retail industry focused on retail trends, Loss Prevention, Asset Protection and IT Security.
Video AI+
Video AI in retail (sometimes known as retail video analytics or AI video intelligence) is the use of computer vision and machine learning algorithms to process live or recorded video feeds from store cameras, turning visual data into automated, actionable business and security insights.
Contributors. Gordon Smith (Market Insights/QA), Sean Colarossi (Content Production/QA), Jimmy Girot (Content Production/Data Analysis)
For questions or feedback, contact us at research@telakus.com.
This work is licensed under CC BY-ND 4.0
