Skip to main content

Snyk UNFILTERED: Inside Our First-Ever Customer Summit

October 28, 2026

0 mins read

What happens when you bring security leaders from some of the world's most forward-thinking organizations into one room, close the doors, and invite them to speak candidly about what's working, what's breaking, and what they’re still working out?

That’s what we set out to find with Snyk UNFILTERED, our inaugural Customer Summit. Held in San Francisco ahead of the launch of Evo Agentic AppSec and the AI Security Summit, Snyk UNFILTERED was designed to feel nothing like a conference.

We brought together a small group of security leaders from our customer community for an afternoon of candid discussions, interactive debates, and peer-to-peer exchanges. With no traditional presentations, product demos, or carefully rehearsed talking points, the summit gave attendees the freedom to share their experiences without attribution. What followed was an open exchange of ideas about the challenges and opportunities shaping the future of AI security. Leaders compared notes on real rollouts, real roadblocks, and real questions they have not yet figured out, with a level of transparency that is hard to find at traditional industry events.

From the growing backlog of vulnerabilities to the governance of autonomous agents, attendees challenged assumptions, compared real-world experiences, and explored questions the industry is still answering. And that's where the most valuable conversations began.

Three panelists sit onstage beneath a presentation slide reading “Where are you on your journey?” at Snyk UNFILTERED.

Five key takeaways that stuck with us

Five themes kept resurfacing over the afternoon, across every table and every discussion:

1. When AI writes the code, who owns the risk?

If an engineer prompts an AI tool to build an agent skill and that skill causes an issue later, who is at fault? The room kept coming back to the importance of guardrails: clear controls and ownership that hold up even when no human typed the line.

2. The AI conversation has grown up

Just a few months ago, the questions were about cost and token usage. Now leadership wants to know the return on investment, what AI is really costing, and where to set budget caps. AI has moved from experiment to line item with scrutiny to match.

3. Finding vulnerabilities is only half the job

Detection is no longer the bottleneck. The questions have moved to how teams fix issues, how fast they can fix them, and how they reduce human error within the loop. Fixes that stall in review, testing, and merge were a shared frustration.

4. Risk can come from anywhere

When everyone on a team has access to AI, vulnerabilities can come from anywhere. Democratizing AI also spreads responsibility, and attendees talked about how to keep everyone accountable without slowing anyone down.

5. Humans can only do so much

Teams are limited by how much people can review and approve. Leaders are open to trusting agents, but only with controls that an agent cannot bypass, and with governance designed for machines rather than adapted from processes built for people.

From UNFILTERED conversations to the future of AI Security

Snyk UNFILTERED was just the beginning of an important week in San Francisco. The challenges discussed behind closed doors at UNFILTERED reflected a broader shift happening across the industry: security teams are moving beyond simply identifying risk toward enabling organizations to build, innovate, and operate safely at AI speed.

The next morning, we launched Evo Agentic AppSec, and the AI Security Summit brought builders, researchers, investors, and CISOs into the same building. The questions from the customer room about fixing at speed and governing agents were carried straight into those conversations. Hearing those challenges directly from customers shapes what we build next.

Group of customer summit attendees posing onstage beneath a neon backdrop reading “UNFILTERED” and “Snyk”

The best conversations don't end when the event does

The leaders in the room are navigating enormous change. They’re balancing innovation with accountability and securing technologies that evolve faster than the processes designed to govern them. What made the afternoon valuable was their willingness to talk openly about what they haven’t solved yet. Spaces like that are rare, and we plan to keep creating them.

Snyk Unfiltered is part of our commitment to bringing our customer community together through candid conversations and shared learning, and this was the first of many.

Want to be part of what's next? Find your next event.

SNYK CUSTOMER EVENTS

Get the most recent on demand and upcoming customer events

Posted in:

Read more

blog feature ai pink
Blog

AI Model Risk Depends on What the Model Can Reach

A model’s risk score is only part of the picture: its impact depends on the data, tools, and actions its application can reach. Learn how security teams can verify that reach and choose targeted controls.