Jazz has won the 2026 Cybersecurity Startup Accelerator run by CrowdStrike and Amazon Web Services, with support from NVIDIA Inception. It was selected after reaching the final round of a programme that drew nearly 1,000 applicants.
The winner was decided after a final pitch session in San Francisco, where Jazz emerged from a group of six finalists. Organisers said 35 startups were admitted to the eight-week programme, which took no equity and included support from specialists at CrowdStrike, AWS, NVIDIA and other cybersecurity industry participants.
Jazz focuses on data loss prevention, or DLP, a long-established area of corporate security aimed at stopping sensitive information from leaving an organisation inappropriately. Its approach uses artificial intelligence to assess how data moves through a business and reduce the number of alerts shown to security teams.
Above Security was named runner-up for its use of AI agents to manage insider risk investigations and turn alerts into fuller investigative accounts.
The judging panel included CrowdStrike founder and chief executive George Kurtz, Amazon chief information security officer of integrated security CJ Moses, NVIDIA senior director of agentic AI and cybersecurity engineering Bartley Richardson, and entrepreneur and investor Robert Herjavec.
The accelerator has become one route for younger cybersecurity companies to gain visibility with larger technology groups and enterprise buyers. Organisers said this year’s applicant pool reflected growing interest in AI-focused cloud security products.
For Jazz, the award raises its profile at a time when security vendors are trying to show that AI can improve older product categories rather than simply add another layer of complexity. Data loss prevention has often been criticised for generating large volumes of policy rules and alerts while offering limited context around actual risk.
Ido Livneh, co-founder and chief executive of Jazz, said the company had rebuilt DLP with an AI-first design.
“We’re incredibly grateful to be named the winner of the CrowdStrike, AWS, and NVIDIA Cybersecurity Startup Accelerator. DLP has been broken for decades by rule-writing, alert floods, and no real answers. So we rebuilt it from the ground up, from first principles thinking, as an AI-native system to truly understand how data moves through a business. The guidance and support we received through the Accelerator has been invaluable in helping us sharpen the product and validate it against real-world enterprise demands as we scale,” Livneh said.
Industry shift
The programme’s backers presented the result as part of a broader shift in cybersecurity, with startups increasingly building products alongside major cloud and computing platforms from an early stage. That model gives smaller vendors access to infrastructure and technical support, while helping larger groups identify emerging security products that fit their ecosystems.
Daniel Bernard, chief business officer at CrowdStrike, said the route to building a cybersecurity company had changed markedly in recent years.
“The path to becoming a defining cybersecurity company looks nothing like it did just three years ago. Today’s most promising founders are building alongside the platforms that power modern infrastructure from day one. Together with AWS and NVIDIA, this Accelerator is about supporting that innovation early in the startup journey. Jazz stood out for replacing legacy DLP with an AI-native, context-aware approach, using Melody and Context Vault to replace noisy alerts with clear, actionable answers at scale, and demonstrating what it takes to lead in this new era of cybersecurity,” Bernard said.
AWS also stressed the importance of proving that security tools work in live cloud environments rather than in isolation. That reflects pressure on security suppliers to demonstrate operational value to large organisations dealing with growing data volumes and increasingly automated threats.
“Security innovation only matters if it performs in real-world cloud environments. This accelerator connects founders directly with the infrastructure, operational expectations, and scale enterprises demand. What set Jazz apart was its ability to investigate signals with full business context and surface only meaningful incidents with minimal analyst overhead, reflecting the standards modern security teams expect,” Moses said.
AI focus
NVIDIA pointed to the growing role of agentic AI in cybersecurity, where systems are designed to make judgements and act with a degree of autonomy. The idea has gained traction across the sector as vendors seek to use AI not only for detection, but also for triage, investigation and response.
“The shift to agentic AI is transforming cybersecurity, requiring autonomous systems that can perceive, reason, and act against increasingly sophisticated threats. This Accelerator equips the next generation of security platforms with the high-performance compute and foundational frameworks needed to build secure, scalable AI systems. Jazz distinguished itself through its agentic investigator, Melody, and its ability to analyze multi-dimensional context across data, systems, people, and business to drive precise, in-context prevention decisions – distilling millions of events into a small set of actionable insights,” Richardson said.
