Pwn2Own Berlin 2026 rewards $1.3M in zero-days across Windows, Linux, VMware, and AI products
Security researchers demonstrated previously unknown exploits at Pwn2Own Berlin 2026, earning $1.3 million in total bounties across multiple platforms including Windows, Linux, VMware, Nvidia, and AI systems. The event highlights an active market for zero-day vulnerabilities and signals emerging attack surface in AI products.
Affected
Pwn2Own Berlin 2026 demonstrated a continued robust market for zero-day vulnerabilities across multiple product categories. The $1.3 million payout represents significant researcher investment in discovering and developing exploits, with the diversity of targeted platforms indicating that vulnerability research has expanded well beyond traditional operating systems into hypervisors and AI infrastructure.
The inclusion of Nvidia and unspecified AI products as targets marks a notable shift in the competitive vulnerability research landscape. As organisations increasingly deploy AI systems in production environments, the security properties of these systems remain underexplored. Researchers are evidently recognising both the technical challenge and the financial incentive to find exploits in this emerging attack surface. The specific vulnerability types and exploitation techniques remain undisclosed at present, limiting technical assessment.
From a defender perspective, the involvement of VMware is particularly significant given its prevalence in enterprise virtualisation infrastructure. Pwn2Own exploits typically demonstrate novel attack chains that bypass multiple security boundaries. The breadth of platforms targeted suggests researchers are focusing on depth of compromise rather than breadth of affected systems.
The pricing signals matter. High bounties attract sophisticated researchers and indicate that vendors are taking vulnerability disclosure seriously. However, the existence of novel exploits for mature platforms like Windows and Linux suggests that defensive patch cycles and security hardening efforts remain incomplete across many organisations. Organisations should assume that some zero-days demonstrated at Pwn2Own will eventually be reverse-engineered or independently discovered by threat actors.
Defenders should prioritise inventory visibility across all affected product categories and maintain readiness to apply patches rapidly once the exploitation techniques become public. The timing of public disclosure typically follows a responsible disclosure window, during which targeted patching becomes critical. The demonstrated compromise of multiple platform types value of segmentation and least-privilege access controls to contain the impact of any individual compromise.
Sources