Operant AI Launches CodeInjectionGuard to Stop Real-Time Attacks on Autonomous AI Agents
The announcement follows recent security incidents that highlight how quickly the threat landscape is evolving.
Operant AI has unveiled CodeInjectionGuard, a new capability within its Agent Protector platform designed to detect and block malicious code at runtime, addressing a fast-growing security gap in the age of autonomous AI systems.
The launch comes as enterprises increasingly deploy agentic AI tools capable of independently downloading software, executing commands, and interacting with live infrastructure.
The announcement follows recent security incidents that highlight how quickly the threat landscape is evolving. In one case earlier this year, a developer’s system was compromised by a poisoned version of the LiteLLM library uploaded to PyPI just minutes before being automatically installed by an AI-powered development tool.
The malicious package was able to harvest sensitive credentials, attempt lateral movement across systems, and establish persistence within seconds—without direct human involvement.
Such incidents underscore a critical challenge: AI agents operate at speeds far beyond human oversight, dynamically pulling dependencies and executing code from external sources. Traditional security methods like static analysis and pre-deployment scanning are increasingly ineffective against threats that emerge in real-time.
CodeInjectionGuard aims to close this gap by focusing on runtime protection. The system inspects packages as they are downloaded, monitors shell commands executed by AI agents, and enforces restrictions on access to sensitive files such as SSH keys, cloud credentials, and Kubernetes configurations. It also detects and blocks suspicious behaviors like obfuscated scripts, base64-encoded payloads, and unauthorised dynamic code execution before they can run.
According to Operant AI, the technology could have prevented the LiteLLM supply chain attack by intercepting the malicious package before execution. The company positions CodeInjectionGuard as a new standard for securing AI-driven environments, where threats materialise instantly and require equally rapid response mechanisms.
"Finding vulnerabilities and stopping attacks are fundamentally different problems, and the industry is solving them at very different speeds. AI agents can install packages, execute code, and access sensitive infrastructure in seconds faster than any human reviewer, and faster than any static analysis tool can respond. CodeInjectionGuard was built for this reality: defense at runtime, at the point of execution, where the fight actually happens," Priyanka Tembey, Operant AI CTO and co-founder.
CodeInjectionGuard is now available as part of Operant AI’s Agent Protector platform for development and production environments.
Last month, the San Francisco-based startup announced the launch of its AI Infrastructure Ecosystem Partnership Programme, aimed at embedding real-time security directly into AI inference systems as enterprises scale adoption of autonomous and agent-driven technologies.
