As AI Agents Go Rogue, Straiker Launches the Industry’s First Agentic Kill Switch for Agents Building Agents
Learn how Straiker's Agentic Kill Switch helps enterprises immediately disable compromised AI agents and reduce AI security risk.
One control takes any enterprise coding agent and homegrown agents offline in seconds, with a security practitioner making the call.
MOUNTAIN VIEW, Calif. – Aug. 4, 2026 — Straiker, The Agentic Security Company, today announced the Agentic Kill Switch, giving security teams a single control to take a misbehaving or compromised AI agent offline in seconds. Support is available for all enterprise coding agents and agent builder platforms. The launch comes as AI kill switches move from research into the mainstream. Last week OpenAI disclosed that its models reached remote code execution in a controlled test with Hugging Face, and lawmakers proposed requiring shutdown controls for frontier AI. If frontier AI systems need a kill switch, so do the enterprises who already run agents.
The Agentic Kill Switch ships alongside enhancements to Straiker's agent discovery and autonomous adversarial testing, giving enterprises the ability to find their agents, expose how they can be compromised, and immediately stop them when they act outside their intended role.
“Every enterprise deploying AI agents will eventually face a rogue agent moment. The question is whether they can stop it before the agent causes damage,” said Ankur Shah, CEO and co-founder of Straiker. “We built the kill switch because ‘open a ticket’ is not an incident response plan for software that doesn’t wait for business hours.”
Rogue agents can put the entire enterprise at risk
Over the past year, Straiker STAR Labs has continuously red-teamed enterprise AI agents in realistic, production-like environments across financial services, healthcare, and critical infrastructure. Across those assessments, agents exceeded their intended authority in 85% of successful attack scenarios. In one simulated healthcare scenario, a medical aide agent was persuaded to ignore ventilator readings, resulting in a simulated patient death.
The exposure is broad because agents reach across the enterprise’s core systems, including productivity tools, customer relationship management platforms, software supply chains, and infrastructure. A poisoned Skill, a malicious Model Context Protocol (MCP) server, or a prompt injection buried in a document can cause a trusted agent to execute code, expose sensitive data, or misuse connected tools. For an agent with broad permissions, the attack surface can become the entire enterprise.
The challenge is no longer limited to frontier AI labs. As enterprises deploy agents across code, productivity, and business systems, the ability to immediately disable a compromised or misbehaving agent becomes an operational security requirement. That control belongs in the customer’s hands.
Expose how an agent can fail and stop it when it does
Straiker secures agents on two fronts. Before deployment, Straiker’s Ascend AI autonomously red teams agents using real adversarial techniques to identify paths to remote code execution, data exfiltration, tool manipulation, and prompt injection. Security teams can see exactly how an agent can be compromised and address the weakness before deployment.
“Securing AI agents is a new and fast-moving discipline. Straiker’s Ascend AI gives us a rigorous way to red team our agents against real adversarial techniques, and the Straiker team has been a strong partner as we bring the same rigor to our own AI that we bring to protecting our customers,” said the director of application security at a leading security company.
At runtime, Straiker’s Defend AI monitors every request, tool call, and agent-to-agent interaction for the same attack techniques uncovered during red teaming. When an agent crosses a defined security boundary, the Agentic Kill Switch takes it offline, returning control to the security team.
See the Agentic Kill Switch in action
See it live: Straiker will demonstrate the Agentic Kill Switch at Black Hat USA, Booth 2964, and Ai4, Booth 1403. You can also book a meeting directly with our team.
Learn more: Read the product blog for a deeper look at the Agentic Kill Switch.
About Straiker
Straiker is The Agentic Security Company. Its agentic-native platform discovers, tests, and protects the AI agents enterprises run and build, securing the fastest-growing workforce in the enterprise. Straiker is trusted by Fortune 500 enterprises and frontier AI labs, and backed by $85M in total funding, including a $64M Series A led by Marathon with Bain Capital Ventures and Lightspeed. Learn more at straiker.ai.
Media contact
Shannon Van Every
Force4 Technology Communications
Shannon@force4.co
One control takes any enterprise coding agent and homegrown agents offline in seconds, with a security practitioner making the call.
MOUNTAIN VIEW, Calif. – Aug. 4, 2026 — Straiker, The Agentic Security Company, today announced the Agentic Kill Switch, giving security teams a single control to take a misbehaving or compromised AI agent offline in seconds. Support is available for all enterprise coding agents and agent builder platforms. The launch comes as AI kill switches move from research into the mainstream. Last week OpenAI disclosed that its models reached remote code execution in a controlled test with Hugging Face, and lawmakers proposed requiring shutdown controls for frontier AI. If frontier AI systems need a kill switch, so do the enterprises who already run agents.
The Agentic Kill Switch ships alongside enhancements to Straiker's agent discovery and autonomous adversarial testing, giving enterprises the ability to find their agents, expose how they can be compromised, and immediately stop them when they act outside their intended role.
“Every enterprise deploying AI agents will eventually face a rogue agent moment. The question is whether they can stop it before the agent causes damage,” said Ankur Shah, CEO and co-founder of Straiker. “We built the kill switch because ‘open a ticket’ is not an incident response plan for software that doesn’t wait for business hours.”
Rogue agents can put the entire enterprise at risk
Over the past year, Straiker STAR Labs has continuously red-teamed enterprise AI agents in realistic, production-like environments across financial services, healthcare, and critical infrastructure. Across those assessments, agents exceeded their intended authority in 85% of successful attack scenarios. In one simulated healthcare scenario, a medical aide agent was persuaded to ignore ventilator readings, resulting in a simulated patient death.
The exposure is broad because agents reach across the enterprise’s core systems, including productivity tools, customer relationship management platforms, software supply chains, and infrastructure. A poisoned Skill, a malicious Model Context Protocol (MCP) server, or a prompt injection buried in a document can cause a trusted agent to execute code, expose sensitive data, or misuse connected tools. For an agent with broad permissions, the attack surface can become the entire enterprise.
The challenge is no longer limited to frontier AI labs. As enterprises deploy agents across code, productivity, and business systems, the ability to immediately disable a compromised or misbehaving agent becomes an operational security requirement. That control belongs in the customer’s hands.
Expose how an agent can fail and stop it when it does
Straiker secures agents on two fronts. Before deployment, Straiker’s Ascend AI autonomously red teams agents using real adversarial techniques to identify paths to remote code execution, data exfiltration, tool manipulation, and prompt injection. Security teams can see exactly how an agent can be compromised and address the weakness before deployment.
“Securing AI agents is a new and fast-moving discipline. Straiker’s Ascend AI gives us a rigorous way to red team our agents against real adversarial techniques, and the Straiker team has been a strong partner as we bring the same rigor to our own AI that we bring to protecting our customers,” said the director of application security at a leading security company.
At runtime, Straiker’s Defend AI monitors every request, tool call, and agent-to-agent interaction for the same attack techniques uncovered during red teaming. When an agent crosses a defined security boundary, the Agentic Kill Switch takes it offline, returning control to the security team.
See the Agentic Kill Switch in action
See it live: Straiker will demonstrate the Agentic Kill Switch at Black Hat USA, Booth 2964, and Ai4, Booth 1403. You can also book a meeting directly with our team.
Learn more: Read the product blog for a deeper look at the Agentic Kill Switch.
About Straiker
Straiker is The Agentic Security Company. Its agentic-native platform discovers, tests, and protects the AI agents enterprises run and build, securing the fastest-growing workforce in the enterprise. Straiker is trusted by Fortune 500 enterprises and frontier AI labs, and backed by $85M in total funding, including a $64M Series A led by Marathon with Bain Capital Ventures and Lightspeed. Learn more at straiker.ai.
Media contact
Shannon Van Every
Force4 Technology Communications
Shannon@force4.co
