DeepKeep launches PII guardrail to secure enterprise AI

DeepKeep has launched a new guardrail for personally identifiable information that scans and filters personal data in prompts and responses across enterprise AI systems.

The PII guardrail sits within DeepKeep's existing AI Firewall, which contains more than 60 pre-built controls inside its AI security platform. The company targets large organisations that run generative AI and other machine learning tools across business functions.

DeepKeep said the new feature detects, redacts, and blocks personal data inside AI workflows. It focuses on data that appears in prompts from employees and in generated outputs from models. The company said this aims to reduce the risk of unintended data leakage as AI use expands in production environments.

Enterprise adoption of generative AI has raised concerns about how sensitive data flows into and out of models. Business users often paste internal information into chat-style interfaces, while models can surface or infer personal details from training data or integrated databases.

DeepKeep cited industry research that it said shows 69% of business leaders view AI data privacy as a top concern this year. The company said PII remains a prime target for cyberattacks and accidental exposure, and that many existing controls inside organisations do not extend into model interactions and outputs.

Regulatory pressure

The PII guardrail is designed to sit as a security layer across existing AI operations. DeepKeep said it supports compliance with major data protection laws including Europe's GDPR, the California Consumer Privacy Act, Brazil's LGPD and South Korea's PIPA. It said the product focuses on detecting sensitive data in context, rather than only in fixed formats.

Regulators in multiple jurisdictions have increased scrutiny of AI deployments that process personal data. Data protection authorities have issued guidance on training data, retention periods and individual rights requests, and have opened investigations into how companies handle personal information inside AI workflows.

DeepKeep said its system assesses PII inside prompts and responses before they reach external models or internal applications. The guardrail then redacts or blocks that information based on configured policies. The company said this approach aims to reduce the risk of non-compliant processing and reputational damage from breaches.

The product covers categories such as names, contact details, government identifiers, financial information and other sensitive attributes, according to the company. It said customers can adapt detection and redaction policies to their own regulatory and business requirements.

Context-aware detection

DeepKeep compared the new guardrail with other tools in the market, including Microsoft's Presidio. It said its product detects a wider range of personal data categories and applies higher contextual precision during classification.

The company said the guardrail uses AI models that identify personal data even when it deviates from standard formats such as typical phone numbers or email addresses. It said the models apply contextual understanding to decide whether data should be redacted or allowed to remain visible.

DeepKeep gave the example of email addresses. The system would block many instances of personal or internal emails. It would allow a generic, publicly available enterprise support email address to pass through when configured accordingly.

Internal testing and benchmark evaluations showed stronger data classification performance than rival tools, DeepKeep said. It reported improved detection accuracy and fewer false alerts, and said this led to smoother compliance workflows for security and data protection teams.

The company said customers can deploy the guardrail across multiple AI applications through a central platform. Security teams can view detection events and policy decisions through a single dashboard, and can adjust thresholds and rules as regulations or business needs change.

Executive view

DeepKeep positions its platform as a way for enterprises to manage risk across the AI lifecycle, including large language models and computer vision systems. It offers products such as an AI Firewall, automated AI red teaming tools and model scanning services.

Yossi Atlevet, Chief Technology Officer at DeepKeep, said demand for stronger privacy controls is rising as AI becomes embedded in day-to-day operations.

"As reliance on AI continues to grow exponentially, enterprises are increasingly reminded that they cannot afford to cut corners on AI security," said Yossi Atlevet, CTO of DeepKeep. "Our PII guardrail allows businesses to reclaim their power over data security, with a faster and more detailed layer of protection, ensuring that privacy and compliance are built into every stage of AI usage. With a continuous, intelligent layer of defense, this guardrail affords enterprises the ability to innovate responsibly and protect their employees' and their customers' data and trust."

The PII guardrail is available immediately within the DeepKeep platform, and existing customers can activate it through their dashboards.