1Password

Zero-Shot Learning is a podcast about how AI gets built, secured, and deployed. Hosted by Nancy Wang, 1Password CTO, and Dev Tagare, Senior Director of Engineering at Google, it's a builder's view of the architecture and the complex choices it takes to ship with AI. As Chief Product Officer at Vercel, Tom Occhino joined Zero-Shot Learning to discuss how AI is reshaping the developer workflow, from frontend architecture to v0, Vercel's production-ready AI coding assistant. What started as a conversation about AI-assisted development became a case for access control as a design decision, not a security afterthought. How AI changes the developer security model As part of the team that built and shipped React at Facebook, Tom helped replace MVC patterns with a component-based model that changed how an entire generation of engineers reasoned about interfaces. He calls what's happening now with AI-assisted development "a fundamentally different approach to software." Where the earl

May was A&PI Heritage Month, and at 1Password, we're proud to shine a light on the people who bring these perspectives to life in our work and help shape our culture every day. This year, we decided to spotlight Stephanie Cheng, Senior Customer Trainer and a leader within our A&PI Employee Resource Group. With over five years at 1Password, Stephanie has built her career around helping people feel capable, confident, and supported, whether she's onboarding a new customer or creating space for her colleagues to connect and belong. We sat down with her to learn more about her approach to customer education, the community she helps lead, and what A&PI Heritage Month means to her. You’ve spent five and a half years training thousands of customers on something they rely on every day to stay secure. What has that experience taught you about people, about learning, and about what it really takes for something to click? One thing this experience has taught me is that learning really

Note: This blog is a recap of 1Password’s recent webinar, “The unmanaged stack: Governing SaaS apps and AI tools outside SSO.” Head here to watch the complete webinar recording. In the constantly evolving world of enterprise tech, there’s one thing that IT and security teams have always been able to count on: users won’t follow policy if they think it’s standing in the way of their productivity. Case in point: 1Password’s most recent annual report found that 52% of employees have downloaded apps without IT approval. These shadow IT apps typically sit outside a company’s SSO provider , and introduce both unmanaged risk and cost. That governance gap has become more pressing with the growing adoption of AI tools and agents, which introduce new and worsening threats. This issue was the focus of 1Password’s recent webinar, “The unmanaged stack: Governing SaaS apps and AI tools outside SSO.” What is the unmanaged stack? It refers to all of the SaaS apps and AI-based tools that can’t be manag

At 1Password, our Jewish 'Bits Employee Community Group exists to create space for Jewish employees and curious allies alike to connect, learn, and show up authentically. For Jewish Heritage Month this May, we wanted to spotlight Nicole Smith , Staff Project Manager and lead of our Jewish 'Bits ECG. In her four years at 1Password, Nicole has been someone people turn to when the work gets complicated, because she builds trust that makes honest conversations possible. That same instinct to lead with curiosity and create space for real connection shows up across everything she does, from leading complex, cross-functional work to fostering community that enriches how we all experience 1Password together. We sat down with Nicole to talk about what drives her, the projects she's most proud of, and what she hopes her colleagues take a moment to appreciate in honor of Jewish Heritage Month this year. As a Staff Project Manager for our CX organization, you're known for bringing clarity to compl

In our view, The Gartner® Hype Cycle™ for Agentic AI , published in April 2026, contains a passage that we at 1Password feel should be a wakeup call for any organization building an agent program. Gartner states, "In practice, fully autonomous agents are not ready for most enterprise use cases, and human oversight remains essential. Semiautonomous deployments, where there is some human supervision of the work of AI agents, are what enterprises must plan for." That is a clear directive for organizations, and the gap between what it requires and what most current deployments provide should be driving every enterprise AI architecture conversation right now. The Hype Cycle for Agentic AI report states that “Interest in AI agents is significant and accelerating. According to Gartner’s 2026 CIO and Technology Executive Survey, only 17% of organizations have deployed AI agents so far, but 42% expect to do so in the next 12 months, and another 22% within the following year. This is t

Coding agents like Codex are helping developers write, execute, and prepare code for production. Every action that AI coding agents take against a database, an API, or a deployment pipeline requires access to credentials. Today, these credentials typically live in .env files, scripts, or hardcoded in repositories, where they can be easily exfiltrated and are difficult to govern and audit. The shift from AI assistance to AI execution has outpaced how teams manage the secrets needed for execution. 1Password and OpenAI are working together to close this gap. The 1Password Environments MCP Server for Codex makes 1Password the trusted access layer for Codex: credentials are issued just-in-time and scoped to the task, while keeping them outside the model’s context window. Developers get the access they need to build and ship, while secrets stay where they belong. The same integration helps catch secrets at the source. Codex can be prompted to use 1Password and the 1Password MCP to store and

There’s a question we get asked constantly, and it’s the right one to ask: “Can 1Password see the contents of my vault?” The answer is no, and it’s because of how we built the product, not just a promise we’re making. That’s an important distinction, because “we promise” has never been an acceptable answer in this industry. After all, promises get broken, and companies get compromised, acquired, and are under constant attack from threat actors. 1Password’s commitment to our security principles is genuine, but what matters more is how we’ve built that commitment into our product and architecture, and the transparency we back it up with with our security white paper . So here’s the precise answer: The way 1Password is built means that we are incapable, on a technical level, of decrypting and reading your vault contents. We're not policy-prevented or contractually restricted; we are technically incapable . This post explains what that means, why we built it this way, and what the real tra

Zero-Shot Learning is a podcast for AI builders, hosted by Nancy Wang, Chief Technology Officer at 1Password, and Dev Tagare, Senior Director and Head of Engineering for Gemini Enterprise & Business at Google. Together, they’ve built and scaled AI systems at the infrastructure and product layers and bring a builder's perspective to every conversation. The name, zero-shot learning, is an AI concept about applying existing knowledge to new tasks without specific training. For this show, it’s also the premise. Each episode features in-depth discussions with the people building today’s most advanced AI systems, unpacking how agentic architectures are designed, and how teams handle the realities of shipping systems into production. Authentication is built on the assumption that identity can be verified once and trusted for a specified period. Over time, the security industry has gotten very good at validating that trust through a chain of identity providers, certificates, and infrastruc

Design system work follows a well-defined loop: read the ticket, check the Figma spec, find the right component primitives, apply the right tokens, write the Storybook stories, run the tests, open the PR. The steps are consistent enough that when we looked at our design system backlog, we didn't just see a list of tasks; we saw a set of instructions waiting to be executed. So we set an agent loose on the loop. At first, it was a semi-hot mess. But then we gave it the right context, and boom, it has completely changed how we improve our Design System. Here’s our approach on what we did and what we learned. Why we started with our design system Every team considering agentic coding faces the same question of where to begin. The tempting answer is your largest codebase or your most complex feature. The right answer is wherever the work is most well-specified, and the feedback loop is fastest. Our React component library, the web layer of our design system, happened to be both. Conventions

AI coding tools have changed who builds software. The barrier to entry has dropped to the point where a designer, an analyst, or a first-time founder can turn an idea into a working app in an afternoon. That shift is real, and it's accelerating. But every app needs to talk to something. Every API call, database connection, and automated workflow runs on secrets: API keys, tokens, SSH keys, service account credentials. And those secrets have to live somewhere. For most people building with AI tools today, secrets end up in a .env file, a chat message, a script, or a note that will "definitely get cleaned up later." AI coding tools are good at helping you get something working fast, but they tend to suggest the fastest path to a functioning prototype, not the most secure one. The result is real credentials stored in plain text, scattered across machines and codebases, hard to track and easy for threat actors to find when a machine is compromised. This is how credential sprawl s

Today we're releasing the 1Password Device Trust MCP Server, an open-source server that connects your Device Trust data directly to the AI tools your team already uses, like Claude or ChatGPT. It's available now for all customers on Device Trust Connect. As AI agents take on more of the work across your organization, IT and security teams need visibility and control that keeps pace. The Device Trust MCP Server is part of how 1Password is extending that control to the way security teams actually work today, inside AI tools, in plain language, with every action logged and auditable. Once it's running, you can query your entire device fleet without leaving your AI client. Which devices have disk encryption off? Who owns the machines failing compliance checks? How long does it typically take to resolve a specific issue across the fleet? Instead of navigating dashboards or writing custom scripts, you just prompt . What is MCP, and why does it matter? If you use AI tools like Cursor or Claud