At Absolute Security, our mission is to Stop Downtime from impacting your business. When things go wrong, whether it be cyberattacks, bad patches, or user errors, you need your machines to return to a safe and operational state as soon as possible. You must also stop downtime that affects your critical AI agents. Scalable and remote repair are powerful, but AI is driving new attacks, forcing new updates, shipping new code, and introducing new bugs faster than ever. At these speeds, autonomy is the only answer. The risks AI creates are the risks only AI can keep pace with.
Agentic Resilience uses AI to learn how your machines fail and how to fix them before you even see a problem. This is the future of resilience and stopping downtime.
With great power comes great responsibility. We need to empower Resilience agents to perform highly privileged, low-level operations on your machines while at the same time preventing coding agents, business agents, and all our other AI tools from running amok. We need fine-grained control of what each agent can do, where it can go, what it can see, and on whose behalf it acts. Most importantly, we need this level of control at scale. This is exactly what NVIDIA Open Agent Safety Platform provides.
NVIDIA's Open Agent Safety Platform combines NVIDIA OpenShell open-source software for secure agent runtime boundaries with NVIDIA Sentry reference system design, which uses NVIDIA DOCA for out-of-band monitoring and policy enforcement on NVIDIA BlueField-4 DPUs. This gives you both the power and the control you need to deploy Agentic Resilience.
With the launch of NVIDIA Open Agent Safety Platform, we are building the next generation of Agentic Resilience. Agents running on your machines can be empowered to not just detect the problems that cause downtime, but to also fix them. Your IT and security teams will be able to see exactly what actions were performed, and when and where they were done. This gives you the confidence to offload to these agents’ large swaths of the heavy lifting needed to keep your employees working. In turn, this will free your key security and IT team members to perform more critical work, while also slashing recovery times when inevitable bad things happen.
By sitting on top of NVIDIA's OpenShell, coupled with the management and visibility from the new NVIDIA Sentry management plane, together, we can build agents that get their own security policies and their own network access controls. A user’s coding agent shouldn't have the authority to replace device drivers or reinstall parts of the operating system, but a machine’s Resilience Agent needs that power. Now, we can give these agents the rights needed to do their job while also giving you the ability to see every action they perform. Going forward, agents at the edge will be able to ask larger, more powerful and knowledgeable agents hosted in the cloud for additional privileges to perform specific tasks. Then, just like junior IT team members today, these agents can be granted temporary authority as needed. If the edge agents try to exceed their authority, you will be able to see it happening, then higher-level AI systems or human operators can intervene and investigate. The agents get the power they need. You maintain the control and visibility you need.
At Absolute Security, we are excited to support NVIDIA Open Agent Safety Platform, to deliver the control, visibility, and Resilience you need for securing your agentic future. We are working together to ensure our patented persistence-enabled solutions support NVIDIA’s end user devices and customers, to stop downtime for your staff and for your AI agents. Together, NVIDIA and Absolute are leading the way in making autonomous AI agents safe, at scale.
