Pallma AI Raises $1.6M Pre-seed To Secure AI Agents

Enterprise adoption of AI is rapidly accelerating as large language models and autonomous agents transition from pilots to core operational roles. As organisations deploy agentic applications at scale, reliance on systems operating with minimal human oversight is increasing.

As AI agents assume greater responsibility, they introduce new security and risk challenges. Traditional security tools weren’t built for probabilistic, adaptive systems, leaving them exposed to prompt manipulation, data leakage, and over-autonomy risks that can have serious operational and security consequences.

Read Also - Finnish Startup Carbonaide Raises €3.7M To Scale CO₂-Cured Concrete Production

Pallma AI tackles these risks by offering a centralised security, visibility, and control layer for AI agents. Co-founder and CEO Pavlos Mitsoulis says the company is building an AI-native security platform that detects threats in real time and continuously hardens autonomous, AI-powered systems.

The platform integrates with existing enterprise tech stacks, gathering application and system data and applying AI models to monitor, detect, and mitigate risks such as prompt injection, data leakage, and unsafe agent behaviour in real time. Co-founder and CTO Dionysis Varelas added that Pallma pairs real-time threat detection with actionable insights and automated responses, moving beyond passive security.

We develop advanced AI models to identify vulnerabilities before malicious actors can exploit them and recommend necessary fixes, ensuring AI applications remain continuously hardened against emerging threats, Varelas said.

With the fresh capital, Pallma AI will grow its team, speed up product development and scale to meet rising demand for its AI-native enterprise security platform.

About Pallma AI

Pallma AI provides end-to-end security for AI agents throughout their lifecycle. Its platform offers proactive red teaming, real-time threat detection, and automated remediation, ensuring AI-powered applications operate safely, reliably, and securely while mitigating risks such as prompt injection, data exposure and unsafe autonomous behaviour.