Each year Okta Ventures asks its portfolio founders to give their predictions for the year ahead. If 2025 was marked by an awareness in the market of the potential for AI agents then 2026 is the year we will see them come into action executing end to end processes and challenging incumbents along the way with new, more efficient, and more secure workflows.
Read below to see our founders’ predictions for the year ahead.
Eran Barak, CEO
“As we enter 2026, the lines between user and agent, structured and unstructured data, access and action, and among IT environments such as SaaS and GenAI apps, clouds, on-prem servers, endpoints and emails will continue to blur. What matters now isn't just visibility or control in isolation, it's understanding risk in context in real-time.
Identity and data are no longer separate problems. They're part of the same challenge and must be part of the same solution.
Security teams that embrace convergence, invest in adaptive intelligence and enable autonomy with trust will not only reduce risk, they'll unlock a new era of resilience, productivity and innovation.
The future of security isn't more control. It's smarter guidance.”
Zohaib Ahmed, CEO
“Identity Becomes the Frontline of AI Security. Every major AI threat ties back to identity. As deepfake fraud accelerates in 2026, companies treating AI security as optional will become cautionary tales. AI detection isn't a feature anymore, it's infrastructure.”
Vijay Rayapati, CEO
“The Orchestration Inflection for AI Agents: 2026 is when agentic AI stops being a new feature and becomes part of core enterprise infrastructure. The question isn't whether to deploy agents — it's whether your IT architecture can handle them at scale with security and compliance intact.”
Evan Johnson, CEO
“2026 will be the year vertical AI agents disrupt enterprise incumbents: AI has transformed how individuals work, most notably with AI-assisted coding. Enterprise adoption has been slower though. Building AI solutions that genuinely solve complex enterprise problems takes time to understand needs, integrate with systems, and frankly the models needed to catch up to the hype for tool calling and more heavyweight agents. In 2026, we'll see the first wave of vertical AI companies emerge with enough product maturity to pose real threats to big incumbent SaaS vendors."
