Dangerous Data: Why Meaning Is the New Cyber Battleground | Wendy Nather

What happens when the real risk to your organization isn’t that data gets stolen… but that it gets quietly changed?

In this onsite episode from AtlSecCon 2026 (Atlantic Security Conference) in Halifax, Amy Yee sits down with Wendy Nather, Senior Director of Research at 1Password, former CISO, and a long-time cybersecurity leader with experience spanning financial services, government, and industry research.

Following her keynote on Dangerous Data, Wendy shares emerging patterns that challenge how we think about cybersecurity in the age of AI — including the growing importance of data integrity, the risks of AI-generated meaning, and why our instinct to treat AI like it’s human can lead us in the wrong direction.

As organizations move quickly to adopt AI, this conversation explores what’s actually changing, what isn’t, and why the fundamentals — from identity to access to understanding what you have — still matter more than ever.

It’s a thoughtful look at how trust is evolving, and what leaders need to pay attention to as the line between data and meaning continues to blur.

🎧 Chapters

00:00 Introduction & AtlSecCon context
01:00 Dangerous Data — keynote overview
02:00 What’s changing (and not) with AI
03:20 Toxic anthropomorphism explained
05:00 Why we instinctively trust AI
06:30 When AI gets it wrong (and why)
08:00 Integrity attacks vs traditional cyber threats
09:30 The challenge of proving a negative (breaches & claims)
11:00 Trusting data inside organizations
12:30 The cognitive load of AI
14:00 Weaponization of semantics
15:30 What actually works in cybersecurity?
17:00 AI adoption — where to be cautious
18:30 Mindsets for navigating uncertainty
19:45 Closing reflections