Securing AI at scale: Databricks and Noma close the inference vulnerability gap
1 min read
Summary
Enterprise AI strategies are being hindered by issues with inference, including prompt injection, data leaks and model jailbreaks, according to a VentureBeat report.
Databricks Ventures and Noma Security have joined forces to tackle these issues head on, bolstered by a $32m Series A funding round.
The partnership will incorporate real-time threat analytics, inference-layer protections and AI red teaming directly into businesses’ workflows.
Niv Braun, CEO of Noma Security, said securing AI inference in real-time was crucial, with 80% of AI incidents predicted to arise from internal misuse by 2026.
“Our joint approach enables organisations to accelerate their AI ambitions safely and confidently,” Braun added.