AI PHYSEC TODAY

🔍 Episode Title: ONVIF AI StandardsGuest: Leo Levit, Chairman of ONVIFHost: Don Morron | AI PHYSEC TODAYThe security industry is entering a new chapter and ONVIF is once again at the center of it. In this episode, we break down how the world’s leading standards body for physical security is tackling artificial intelligence and why AI metadata, digital trust, and cross-vendor interoperability will define the next decade of security technology.If you’ve ever wondered what AI standards will actually look like, why they matter, or how ONVIF will influence multi sensor and multi vendor environments, this is your episode.⸻🔥 In This Episode We Cover:• What ONVIF really is and why it was created back in 2008How a common language for IP video solved chaos in the early days of network cameras and VMS platforms.• The new AI problem ONVIF is trying to solveWhy today’s AI systems desperately need a shared way to exchange events, objects, context, and even natural language.• Inside the ONVIF AI Working GroupWhat the team is researching, what problems they’re mapping, and why this matters even before a formal AI profile exists.• Standardizing AI information flowWhat kind of metadata must be shared across devices and applications to make AI useful and interoperable.• Why vendor-locked AI pipelines are a dead endAnd how industry standards free integrators, consultants, and end users from closed ecosystems.• How Model Context Protocol and other emerging AI “languages” could fit into ONVIF’s future workWhere ONVIF sees alignment…and where it sees gaps.• Evolution from Profile M to true AI metadata standardsHow the organization is building on existing analytics work to support far richer, context aware AI.• Digital trust, authenticity, and the threat of deepfakesHow ONVIF is thinking about certifying original media in a world where synthetic content is easy to generate.• Moving from reactive to proactive securityHow standardized AI metadata helps operators prevent incidents instead of replaying footage afterward.• Where AI still falls short todayAccuracy, context, edge compute constraints, market expectations…the real limitations and opportunities.• Cutting through the marketing noiseHow standards bring clarity to what AI solutions can and cannot do.• The future state: multi sensor, multi vendor, AI powered securityWhat a fully interoperable environment could look like in 3 to 5 years.• How integrators, consultants, and end users can prepare nowSteps to stay ahead and avoid being locked out when AI-driven cloud and metadata features become the norm.⸻👍 If you work in physical security, this episode is a must-watch.You will walk away knowing exactly where AI standards are heading and how they will reshape your operations, deployments, and long-term strategy.Subscribe for more AI and physical security insights every week.Leave a comment with your biggest takeaway or question for Leo.

Is the software industry facing a "mass extinction" event?In this episode of AI PHYSEC TODAY, we sit down with Steve Van Till, Founder & CEO of Brivo, to unpack his controversial and transformative thesis: AI is eating software.For decades, Marc Andreessen’s famous "Software is eating the world" mantra defined the tech landscape. But as Steve argues in his recent piece for Security Journal Americas, that era is over. We are now entering a phase where AI-native agents and autonomous code generation are dismantling traditional SaaS moats.What does this mean for the physical security industry?When a small team can clone a major SaaS platform in just 30 hours using AI, the "Build vs. Buy" calculus changes forever. Steve breaks down which parts of the security stack are vulnerable, which "moats" (like field service and regulations) might actually hold, and how integrators can pivot from being software resellers to AI orchestrators.🔥 In this episode, we cover:- The "Mass Extinction" Event: Why traditional SaaS categories are the most vulnerable to AI disruption.- 30-Hour Clones: The shocking reality of how fast AI can replicate complex legacy products.- The Physical Security Moat: Can field technicians and fire codes protect our industry from pure-digital disruption?- Build vs. Buy 2.0: Criteria for deciding when to build your own AI capability versus buying off the shelf.- The Integrator Opportunity: How service providers can ride the wave instead of being displaced by it.- The New Software Layer: If software is "eaten," what replaces it? (Hint: It’s about orchestration and agents).👤 About Our Guest:Steve Van Till is the Founder and CEO of Brivo, the global leader in cloud-based physical access control. A pioneer in the security industry, Steve is also the author of The Five Technological Forces Disrupting Security and serves as Chairman of the SIA Standards Committee.🔔 Subscribe for more insights on the future of AI in Physical Security.#AIEatsSoftware #PhysicalSecurity #SaaS #AI #ArtificialIntelligence #Brivo #SecurityTechnology #PropTech #DigitalTransformation #Innovation #FutureOfWork #AIAgents #TechDisruption #BuildVsBuy #SystemsIntegration

Can we have real-time public safety without sacrificing privacy? This is the critical question facing the physical security industry.The answer isn't just "more AI," it's a new kind of AI.In our latest podcast episode, expert Steve Lindsey, Founding Partner of LVT and I explore the shift to Agentic AI. This isn't your traditional, rule-based analytics. Agentic AI is designed to be proactive and autonomous, using complex reasoning to act on goals.We dive deep into:The "Why": Why agentic AI is a pivotal transformation, moving security from passive monitoring to proactive, real-time deterrence.The "How": How we can identify suspicious behavior (proactive safety) without resorting to invasive biometric tracking or facial recognition (privacy violation).The Balance: Why safety and privacy are not mutually exclusive, and the critical role transparency plays in building public trust.The Future: What new AI governance frameworks are needed to ensure ethical, responsible deployment.This is the future of trusted security. If you're in security operations, building management, or tech innovation, this conversation is for you.Listen to the full episode and let us know what you think!#AgenticAI #PhysicalSecurity #AI #EthicalAI #ResponsibleAI #Privacy #PublicSafety #AIGovernance #ArtificialIntelligence #SecurityTech #Innovation #VideoSurveillance #RiskManagement #BehavioralAnalysis #DigitalTransformation

EP32: Is The Physical Security Industry Ready For AI?

🎙️ AI PHYSEC TODAY – Episode 31 Title: How to Operationalize AI Guest: James Connor, Head of Corporate Engagements at Ambient.ai AI in security is often talked about in the abstract. In this episode, we focus on what it means to operationalize AI—to move beyond pilots and into practical impact. We discuss: Why AI should be seen as augmentation rather than replacement The challenges guard forces face today and how AI can help address them How GSOCs are shifting from “eyes on glass” to intelligence-driven operations Real proof points: reducing false positives, improving response times, and driving ROI The convergence of physical and cybersecurity, and what the next decade may bring James also shares advice for young professionals preparing for a future where AI is part of every security role. #AI #Security #Innovation #FutureOfWork #OperationalAI

EP30: Why Security Leaders Should Be Bullish on AI

EP29: Agents vs Agency

EP28: Real-Time AI

EP27: Augmented Intelligence (AI)

EP26: What is VisionOps?

E25: Autonomous Security Agents

E24: Security Robotics

E23: AI in Retail & Restaurants

0:00 Intro 1:33 Podcast Guest Intro 4:53 Chapter1: Evolution of Smart Buildings & AI 5:08 From Traditional to Smart 10:08 Reactive vs. Proactive Security 21:01 Real Estate Companies & AI 26:00 Chapter 2: The Power of Connected Data & Generative AI 26:10 Breaking Down Data Silos 31:56 Developing Savvy 39:46 Instant Insights for Real Estate Pros 51:00 Chapter 3: Future Trends and Considerations 51: 28 Emerging Technologies in Smart Buildings 55:57 Cohesion’s Forward Strategy 57:49 Advice for Real Estate Professionals58:35 Outro

EP20 AI PHYSEC TODAY Guest: Florian Matusek 0:00 Intro 1:06 Podcast Intro 3:57 Evolution of Security Camera Technology 15:50 Practical Applications and Benefits 23:33 Camera Analytics, Machine Learning or Deep Learning? 29:10 Technical Challenges and Considerations 37:19 Ethical and Privacy Implications 48:11 Future Directions and Innovations

BEARDCAST Mini-Series, S2EP09 AI PHYSEC TODAY Location: San Francisco, CA 0:00 Intro 1:05 Conference Highlight Reel 1:41 BEARDCAST Intro 2:00 Day One TLDR 2:35 Open Source AI 5:37 PHYSEC Inferences - Open Source AI 6:40 Knowledge Graphs 7:41 PHYSEC Inferences - Knowledge Graphs 8:00 On Device LLMs 9:19 PHYSEC Inferences - On Device LLMs 10:44 Relationship Between Machines and People 12:25 PHYSEC Inferences - Machines and People 13:51 AI Networking 14:25 PHYSEC Inferences - AI Networking 16:28 Large Design Models 17:30 PHYSEC Inferences - Large Design Models 17:58 AI Agent Evaluation 18:32 PHYSEC Inferences - AI Agent Evaluation 19:22 Advanced RAG 20:02 PHYSEC Inferences - Advanced RAG 20:25 Make AI Like Software 21:07 PHYSEC Inferences - Make AI Like Software 21:31 AI Legal Issues 22:08 PHYSEC Inferences - AI Legal Issues 23:10 AI AI-Engineer 23:40 PHYSEC Inferences - AI AI-Engineer 24:21 AI Transforms Customer Success 26:48 PHYSEC Inferences - AI Transforms Customer Success 27:17 Day One PHYSEC Inference Summary 28:25 Outro

🎙️ Podcast Episode: AI PHYSEC TODAY - Episode 8, "The AI Operator" 🚨 Revolutionizing Security Operations with AI 🚨 This episode explores the HiveWatch AI Operator, powered by Anthropic’s Claude AI, designed to tackle major challenges in physical security. Here’s a quick breakdown: ➡️ Origins: Built to solve inefficiencies, high turnover, and false alarms in security teams. 📉 False Alarms: Cuts through the noise, focusing only on real threats. 🤖 Claude AI: Advanced AI for handling complex security scenarios. 📋 Custom SOPs: Tailored to fit each organization’s processes seamlessly. 💡 Efficiency Boost: Automates mundane tasks, empowering teams for strategic action. ⚡ Real-Time Response: Reliable, accurate decision-making for critical alarms. 💰 ROI: Saves costs while enhancing operations. 🔮 Future-Proof: Prepares organizations for an ever-evolving threat landscape. 📢 Don’t miss this transformative discussion! Tune in to learn how AI is shaping the future of physical security. #AI #PhysicalSecurity #TechInnovation #FalseAlarms #ClaudeAI #Automation #FutureOfWork #HiveWatch #SecurityTech #PodcastEpisode #AIInnovation #SecurityOps 0:00 Intro 0:42 HiveWatch AI Operator Promo Video 3:09 Podcast Intro 4:21 Origins of Innovation 8:25 Tackling False Alarms 11:11 Strategic Intelligence Overload 16:25 Leveraging Claude AI 21:00 Customizable SOPs 26:37 Enhanced Security Operations 28:43 Operator Efficiency and Turnover 32:05 Real-Time Decision-Making 38:30 Data Security and Privacy 41:26 Success with Early Adopters 42:16 Human-AI Collaboration 48:05 Reducing Costs and Enhancing ROI 53:56 Incident Documentation 56:12 Future-Proofing Security 57:22 HiveWatch’s Vision 58:37 Outro

5 AI News & Inferences For the Physical Security Industry (S2 EP07 BEARDCAST - Mini-series within AI PHYSEC TODAY) Robotics: NVIDIA Omniverse - https://youtu.be/l5M4sqaRd6w?si=weKWJBDl7aUORVKw Boston Dynamics New Spot Acoustics & Other Upgrades - https://youtu.be/0hrYzgP_Lg4?si=OyrTl_P0bSSQDtyQ AGI: OpenAI o1-preview Overview - https://openai.com/index/introducing-openai-o1-preview/ OpenAI o1 vs 4o - https://openai.com/index/learning-to-reason-with-llms/ OpenAI pitches nuclear power to Whitehouse - https://www.energyconnects.com/news/renewables/2024/september/openai-pitched-white-house-on-unprecedented-data-center-buildout/ OpenAI Agents Coming in 2025 - https://www.tomsguide.com/ai/chatgpt/the-agents-are-coming-openai-confirms-ai-will-work-without-humans-in-2025 Sam Altman's Age of Intelligence manifesto - https://ia.samaltman.com/ Google NotebookLM - https://notebooklm.google.com/?pli=1 Microsoft Copilot Updates - https://www.microsoft.com/en-us/microsoft-365/blog/2024/09/16/microsoft-365-copilot-wave-2-pages-python-in-excel-and-agents/ Microsoft AI Correction - https://www.theverge.com/2024/9/24/24253452/microsoft-correction-ai-safety-tool-fix-errors Google Deepmind AlphaChip - https://deepmind.google/discover/blog/how-alphachip-transformed-computer-chip-design/ Spatial Computing: Meta Orion AR Glasses - https://youtu.be/pPLWIL64sgQ?si=8xLDgr1fF-aKqzIj World Labs Spatial AI - https://www.worldlabs.ai/about Fei Fei Li on Spatial Intelligence - https://medium.com/@jckapadia003/world-labs-takes-off-fei-fei-lis-new-venture-a413dda29e21 Regulation: CA AI Bill Vetoed - https://www.gov.ca.gov/wp-content/uploads/2024/09/SB-1047-Veto-Message.pdf Industry Insights: Veesion Shoplifting AI - https://veesion.io/en/ University of Maryland Invisible Sweater (Cloak) - https://www.cs.umd.edu/~tomg/project/invisible/ 0:00 Intro 0:55 Robotics 0:57 Investment Outlook 5:08 Nvidia Omniverse 8:21 Boston Dynamics Spot Upgrades 11:52 Robotics Physical Security Inferences 14:46 AGI 15:10 OpenAI GPT-o1 21:47 OpenAI GPT-4o vs GPT-o1 26:05 OpenAI Advanced Voice Mode [DEMO] 29:23 Sam Altman, Nuclear Power, & US Congress 31:14 Agents Coming in 2025 32:28 The Age of Intelligence by Sam Altman 33:38 ChatGPT Canvas [DEMO] 38:19 Google NotebookLM [DEMO] 41:41 Microsoft Copilot Updates 48:03 Microsoft AI Correction 49:05 Google Deepmind's AlphaChip 51:44 AGI Physical Security Inferences 58:48 Spatial Computing 58:51 Meta Orion AR Glasses 1:03:14 World Labs' AI Model 1:06:20 Fei Fei Li on Spatial Intelligence 1:07:39 Spatial Computing Physical Security Inferences 1:11:13 Regulation 1:11:17 CA AI Bill Vetoed 1:13:29 Regulation Physical Security Inferences 1:15:54 Industry Insights 1:16:00 Shoplifting AI 1:18:44 The Invisible Cloak 1:20:10 Industry Insights Physical Security Inferences 1:22:12 Outro

0:00 Beardcast Intro 0:39 Episode Introduction 2:15 Humanoid Robots3:22 1X Neo Beta6:15 Humanoid Robot Inferences 8:10 Artificial General Intelligence (AGI)10:50 OpenAI Strawberry14:38 Altera Project SID16:45 Fully AI Generated DOOM Video Game18:23 AGI Inferences 19:57 Spatial Computing22:09 Midjourney Expands to Hardware23:40 Spatial Computing Inferences 24:32 AI Regulation25:23 California's AI Safety Bill: Companies For & Against27:16 AI Regulation Inferences 29:18 Industry Specific29:22 Wyze AI Search31:30 Risk of Deep Fakes 34:32 Conclusion

Daniel Reichman, Ph.D., CEO & Co-Founder of Ai-RGUS dives into "Upskill with AI".