Oso

Nobody can tell the difference between you doing something and your AI agent doing it. That's the whole problem with agent identity right now, and it's the same access problem security teams never solved for humans. Graham Neray, Co-Founder and CEO Oso spoke to Ashish and Caleb about why agents go straight for production credentials. The reason agents are dangerous is the same reason they're useful: you hook them up to production systems and real data. Most of those systems are badly over-permissioned, and everyone already knows it. We never solved least privilege for slow, non-automatic humans. Access scoped to exactly what you need, only when you need it, only for as long as you need it. Agents just expose that gap at machine speed. Save this for the next time someone pitches you an agent that needs full prod access on day one. Follow AI Security Podcast for practitioner takes on securing AI systems. #aisecurity #cloudsecurity #identitysecurity #leastprivilege #aisecuritypodcast

"Passwords are the keys to the kingdom. You need to go to someone that's trusted." - Graham Neray, Co-founder and CEO of Oso on why the team started using 1Password from day one. With the rise of AI agents, Graham says credential security is “table stakes.” We agree. Credential sprawl doesn't announce itself. It accumulates quietly across SaaS apps, shared logins, AI tools, and accounts created to get work done. That’s the problem 1Password is built to solve, for every employee, agent, and machine identity in your organization. #IdentitySecurity #CredentialSecurity #Startups #Cybersecurity #AgenticAI #1Password

"I thought I was being so clever. We don't need to build blocking, no one's ever gonna use it." Then every buyer asked for it anyway. So they built it. And almost nobody turns it on. Graham Neray of Oso spoke to Ashish and Caleb about why the most demanded control in agent security sits switched off and what teams actually run instead. Human in the loop. Security teams still trust their engineers to apply judgment. They just want someone looking closely at what the agent is about to do, before it does it. Blocking gets you through the security review. Human in the loop is what engineering teams actually leave on. Caleb called it the always untold story of security, the feature everyone demands and no one uses. Follow AI Security Podcast for practitioner takes on securing agents. #aisecurity #aisecuritypodcast

Why are security leaders terrified of connecting AI agents to production data? Because unlike humans, AI agents don't apply judgment, and they operate at machine speed, meaning they can relentlessly hunt down production credentials and do catastrophic damage before a human analyst even blinks. In this episode, Ashish and Caleb sit down with Graham Neray, CEO of Oso, to tackle the massive, unsolved problem of AuthZ (Authorization) for autonomous AI. We explore why the industry's reliance on static, over-permissioned human identities is a recipe for disaster when applied to tools like Claude Code and Notion Agents. Graham explains the dangerous pitfalls of allowing agents to adopt the permissions of their human operators (privilege escalation), versus the complexity of assigning agents their own unique service accounts. The conversation dives deep into the fragmented agent security market. Should you deploy a browser extension, an endpoint sensor, or an edge proxy?. Learn why blocking destructive actions is a flawed approach (because agents *need* to destroy things to work), and why the future of AI AuthZ requires dynamic, data-level policies and continuous "human in the loop" validation. #aisecurity #aisecuritypodcast