Barracuda AI Data Security blocks sensitive data from ChatGPT

Real-time blocking for AI data leakage

Barracuda Networks launched Barracuda AI Data Security on September 22, a product designed to monitor and control what company data employees send to artificial intelligence tools. The release addresses a governance gap created when workers adopt chatbots faster than organizations can establish usage policies.

The product runs on Barracuda's BarracudaONE platform and comes with default sensitive-data blocking active from deployment. Detection is handled by Barracuda IQ, which applies policy across more than 1,300 generative AI services including ChatGPT, Microsoft Copilot, and Gemini. Prebuilt detectors scan prompts and file uploads for customer records, passwords, financial data, and proprietary source code, blocking them before they reach external models.

Beyond data blocking, the engine screens for prompt injection attempts, jailbreak techniques, hate speech, and employer-defined prohibited topics. Barracuda said its detection aligns with the OWASP Top 10 for large language model applications.

Built for smaller organizations and MSPs

Barracuda is explicitly targeting smaller companies and the managed service providers that support them. Neal Bradbury, chief product officer at Barracuda, said most governance tooling in this category is built for large enterprises, leaving smaller buyers with skill requirements they cannot meet.

The product includes firewall-style rule screens for administrators, with Barracuda's Bailey assistant available to guide custom rule creation. Policies can be configured per team, per tool, and per data type. Every AI interaction is logged in an audit trail that exports in a single click for auditors and insurers.

Managed service providers receive multitenant management across customer accounts, with pricing bundled into existing service packages. Geoff Turner, chief executive of managed service provider Elevate Technology Group, said customer AI policy "only goes so far" and that his firm needs controls that are "simple, scalable and built for real-world environments."

Why it matters

Barracuda's own survey of 2,000 IT and security decision-makers found that 49% of chief information officers at companies with 500 to 1,000 employees said their teams lack the skills to secure and govern AI already in use. The figure across all respondents was 38%. This skills gap creates real exposure as employees increasingly use AI tools without oversight, potentially leaking intellectual property, customer data, or regulated information to external systems. By offering controls that don't require deep AI security expertise, Barracuda is addressing a market segment that has been largely underserved by enterprise-focused solutions.

Barracuda AI Data Security becomes available in October and is included at no additional cost in Barracuda SecureEdge Premium Access. The company released generative AI risk visibility in BarracudaONE in March at no extra charge, bringing shadow AI detection and risk scoring to existing customers. Bradbury described the new product as the first of several planned additions to the company's AI security line.

Details were first reported by SiliconANGLE.