PREDICTIONS
CONTINUED
Bill Conner , president and CEO, Jitterbit If there is an AI bubble, it will only burst for the people building on hype instead of solid foundations. The ones who’ll come out on top in 2026 are those investing in strong connections between their data, their systems and their people.
The gap between hype and real, long- term impact ultimately comes down to execution. When your data moves smoothly and your systems can actually talk to each other, AI stops being another buzzword and it starts making a real difference. If there is an AI bubble and it does burst for those building on hype, it won’t signal the end of progress, but rather a shift toward organisations that prioritise substance, integration and sustainable innovation. n
AI can’t do anything meaningful if it’s stuck in disconnected silos. The companies winning right now, and who will stay winning in 2026, are the ones using integration and automation to turn accountable AI from a shiny idea into actual business results.
Bill Conner
jitterbit.com
Jamie Moles , senior technical manager, ExtraHop AI agents are increasingly being
gets in and identifying suspicious behaviours early. Businesses need to treat identity as a living, shifting system. In addition, legacy systems will become beacons for threat actors. Everyone knows they’re fragile, everyone’s scared
integrated into business workflows and already becoming a new attack surface as most organisations still have no idea how to govern them. If an attacker compromises one, they’ll get all the access without any of the scrutiny. In 2026, these agents need to be treated like high-risk identities and managed like a normal employee: monitored, restricted and continuously verified. Blind trust in opaque systems is how breaches happen and AI governance will be even more key in 2026. Identity-based attacks will continue to be a focal threat to enterprises in 2026, especially as cloud and third-party integrations multiply. Even with perfect patching and verification tools, a single convincing phishing attack can undo everything. The onus then will be on the internal east–west network traffic – watching what happens after someone
to touch them, and that hesitation is exactly what criminals exploit.
Modernisation is now imperative to risk reduction, it’s not just a nice-to-have. The systems you’re too nervous to change are the ones threat actors will target first. But the human element will remain the weak point no matter how much automation we add. People still read MFA codes aloud, approve unfamiliar access requests, and get pressured into quick decisions. And as agentic tools take over routine tasks, humans can become even easier to trick. Regular reinforcement and removing decision pressure will be the key factors to managing this risk. n
Jamie Moles
extrahop.com
36
Powered by FlippingBook