SPLXAI receives $7 million in seed rounds to deal with growing security threats in proxy AI systems

In an important step towards maintaining the future of AI, Splxai, a pioneer in offensive security against broker AI, raised $7 million in seed funding. The round is led by Launchub Ventures, with strategic participation from Rain Capital, Inovo, Runtime Ventures, DNV Ventures and South Central Ventures. New capital will accelerate SPLXAI PlatformDesigned to protect organizations that deploy advanced AI agents and applications.

As companies increasingly integrate AI into daily operations, the threat landscape is developing rapidly. By 2028, 33% of enterprise applications are expected to combine proxy AI-AI systems that are capable of automatic decision-making and complex task execution. However, this shift has brought about a greatly expanded attack surface that traditional cybersecurity tools cannot handle.

“Massive deployment of AI agents introduces significant complexity,” explain Kristian Kamber,,,,, CEO and Co-founder of SPLXAI. “In this environment, manual testing is not feasible. Our platform is the only scalable solution to ensure proxy AI.”

What is proxy AI and why is security risk?

Unlike conventional AI assistants that respond directly to prompts, Admit AI refers to a system that can automatically perform multi-step tasks. Think of AI agents that can schedule meetings, book trips, or manage workflows without the need for continuous human input. This autonomy, while strong, introduces serious risks including timely injections, subject responses, context leaks, and AI hallucinations (false or misleading outputs).

Furthermore, most existing protections (such as AI guardrails) are reactive, well-trained, usually well-trained, resulting in excessively restrictive behavior or dangerous allowance. That’s where Splxai intervenes.

SPLXAI platform: Red team for large-scale AI

The SPLXAI platform provides Genai Systems with a fully automatic red team, enabling enterprises to conduct continuous real-time penetration testing in AI-driven workflows. It simulates complex adversarial attacks (imitating the real world, highly skilled attackers), including text, images, voice and even documents.

Some great features include:

• Dynamic risk analysis: Continuously detect AI applications to detect vulnerabilities and provide actionable insights.

• Domain-specific pentagons: Tailor-made unique use case testing for each organization – from finance to customer service.

• CI/CD pipeline integration: Embed security directly into the development process to catch vulnerabilities before production.

• Compliance Mapping: Automatic evaluation of alignment with frameworks such as NIST AI, OWASP LLM TOP 10, EU AI ACT and ISO 42001.

This proactive approach is already attracting people. Clients include KPMG, Infobip, Brand Engagement Network and Glean. Since its launch in August 2024, the company has reported a quarter-quarter growth of 127%.

Investors’ Vision to Support AI Security

emissionUB Ventures’ General Partner Stan SirakovNow, he now joins the SPLXAI board of directors, highlighting the need to expand AI security solutions: “With proxy AI becoming the norm, so is its potential for abuse. SPLXAI is the only provider that plans to manage that risk on a large scale.”

Rain capital PhD Chenxi Wang Responding to this sentiment, highlighting the importance of the automatic red team to the AI ​​system during infancy: “SPLXAI’s expertise and technology position it as a core player in ensuring Genai. Manual testing doesn’t cut it anymore.”

New additions enhance the team

In addition to funding, SPLXAI announced two strategic employees:

• Stan Sirakov (Start the company) Join Board of Directors.

• Sandy Dunnformer brand participating in the network CISO, intervenes Chief Information Security Officer Leading the company Governance, Risk and Compliance (GRC) initiative.

Cutting-edge tools: Agent radar and real-time repair

In addition to the core platform, SPLXAI recently launched the proxy radar, an open source tool that maps dependencies in the proxy workflow, determines weak links and analyzes surface security gaps through static code.

Meanwhile, their remediation engines provide an automated way to generate hardened system prompts, reducing attack surfaces by 80%, increasing the speed of rapid leak prevention by 97%, and reducing engineering efforts to 95%. These system prompts are critical to shaping AI behavior and can become a primary security responsibility if exposed or poorly designed.

Simulate real-world threats in more than 20 languages

SPLXAI also supports multilingual security testing, making it a global solution for enterprise AI security. The platform simulates malicious tips of adversarial and benign user types, helping organizations discover such as:

• Context leak (Accidentally disclosed sensitive data)

• Social Engineering attack

• Prompt injection and prison Break technology

• Toxic or prejudiced output

All of this is delivered with minimal false positives thanks to Splxai’s unique AI Red Team Intelligence.

Looking to the future: The future of security AI

As enterprise competition integrates AI into everything from customer service to product development, the need for real-time AI security has never been greater. SPLXAI is leading the charges to ensure that the AI ​​system is not only powerful, but also trustworthy, secure and compliant.

“Our mission is to ensure Genai-powered applications,” Camber added. “Our platform enables organizations to move quickly without breaking things or damaging trust.”

With its new capital and momentum, SPLXAI is expected to become the foundational layer in the AI ​​security stack in the next few years.