RSAC Reflections: Data Is the Foundation for Successful AI

Walking through the Moscone Center at the RSA Conference last week, something felt different. AI was everywhere of course, but the conversation has shifted from the previous two years. Technology leaders aren’t just piloting Chatbots any more, they are looking for ways to actually deliver on the promise of AI. They are under immense pressure and understand that the traditional approaches aren’t effective. Conversations kept coming back to this: You cannot deploy AI confidently without first solving for data security. More specifically, here are the key trends that I observed when speaking with customers, peers and industry leaders: 

Security Teams Enable AI Innovation

Security teams have a legacy of being the department of no or the department of slow. The pressure is no longer just to "allow" AI, but to enable it. The budget has been allocated and the successful security leaders I met with aren’t repeating the sins of the past. They aren’t gatekeepers; they are the ones providing the guardrails that allow the business to accelerate. By securing the data layer first, they’ve become the engine of AI innovation. Like Mario power sliding through sharp corners instead of just applying the emergency brake,  these maneuvers aren’t just for show; they’re necessary because we aren't just talking to bots anymore—we're deploying agents. This innovation is critical as there will be more accountability for the resources the “business” has applied to AI this year. 

AI Innovation Velocity Challenges

It has become more than a full-time job just to keep track of the weekly—and sometimes daily—advancements in artificial intelligence. The innovation cycle is moving so fast that I heard several leaders joke that you practically need to build your own agentic research agent just to ingest, assess and summarize the constant stream of new information. Leaders are looking for a partner to help them keep pace with the velocity of change. 

Data is the Lighthouse For Your AI

In an AI world, securing data and understanding intent become the foundation of the entire security program. Data context will be a key component of AI context engineering. I tested a lighthouse analogy out last week which resonated with many folks. In a sea of change and turmoil you need a focus, something to guide you through the storm to safety. Models will change, tools will change, startups will come and go, but everything will still have to access the data, so in the midst of all this unprecedented change, data will be the lighthouse that keeps your AI deployments from running aground. 

Agentic AI is Real

2026 is officially the year we moved from asking chatbots questions to letting agents take actions. We saw this "consumerization of AI" start early in the year with tools like Open Claw, where people began using agents for personal productivity—integrating APIs and automating life tasks without writing a single line of code. These consumer capabilities are driving the enterprise to consider agents and the art of the possible

The number of conversations I’ve had around agentic use cases has significantly increased this year. In the enterprise, these agents are only as good as the data they can access. Legacy security tools simply aren't built for this autonomous world. For an agent to be effective, data context has to be part of every decision it makes. We are moving toward a model where deep data intelligence is required to orchestrate an agent’s activities

Platform Consolidation Trend

Walking the convention floor, a CISO friend asked me how many of these vendors would be around next year? Buyers are officially over "death by a thousand cuts." No one wants to manage fifty different AI security vendors, especially when those vendors might not exist in six months. That’s why “the platform” was the final trend from last week. The platform is how you enable the outcomes we’ve been discussing. We had one customer say they’d prefer to wait a few months for Cyera to build a capability than go with a stand alone player. A good example of this is AI browser run time protection. There have been companies only focused on this feature, but now it has become a capability within broader platforms. Cyera announced our feature last week. 

Final Thoughts

I left RSAC excited about the future. You cannot deploy AI confidently without first solving data security. With our data centric approach, I’m excited about Cyera’s ability to help every company realize the value of their AI initiatives. To see how Cyera can help your organization secure its data and drive AI innovation, book a demo today at Cyera.com.