Sentra Data Security for AI Agents protects AI-powered assistants

Sentra launched Data Security for AI Agents solution, specifically designed to address the emerging challenges associated with proliferating AI assistants and empower large enterprises to embrace AI innovation securely and responsibly.

With the solution, Sentra also announced platform support for Agent toolkits including Microsoft Copilot Studio, Amazon Bedrock, and OpenAI ChatGPT Enterprise.

Agentic AI holds immense promise to streamline business processes. However, the independent nature of AI agents also introduces new risks of unintended sensitive data access or disclosure. According to a 2024 survey by EY, 80% of respondents worry about AI’s role in cyber attacks.

The Data Security for AI Agents solution builds on Sentra’s rich data security posture management (DSPM) heritage and ensures that AI adoption remains secure and compliant while fostering responsible data usage. Data Security for AI Agents safeguards sensitive information within increasingly autonomous AI workflows, giving security teams peace of mind.

“AI agents help employees and businesses work more thoroughly and efficiently by unlocking creativity, streamlining processes, and providing data-driven insights”, said Yoav Regev, CEO of Sentra. “Organizations must ensure AI agents are used securely, protecting the privacy and integrity of the underlying data on which they depend. Sentra solves for this — providing visibility at the union of agent utilization and sensitive data — enabling teams to facilitate responsible AI application deployment.”

This product launch follows Sentra’s $50 million series B funding round, which is fueling product innovation and investments in securing sensitive data across emerging AI environments. The new AI security capabilities are a key milestone in realizing that vision.

Comprehensive protection for AI copilots and knowledge bases

Sentra’s innovative approach to securing AI copilot agents and their underlying knowledge sources delivers several key capabilities:

• Stack inventory for AI copilots provides automatic discovery and identification of AI agents and their models, the knowledge bases they connect to, along with what sensitive data agents have access to, the users they interact with while running in the environment, and the risk that this creates.
• Data access controls for AI Agents ensure that AI chatbots and assistants operate within defined user roles and permissions when accessing enterprise data sources. By applying intelligent labeling and stringent access controls to files within data sources before they are utilized by AI Agents, Sentra enforces identity-aware policies, guaranteeing that AI-generated responses never reveal unauthorized information.
• Real-time AI data protection provides continuous monitoring of AI agent interactions and remediation recommendations to swiftly detect and resolve instances of unauthorized data access or potential data leakage from malicious actors, helping to stop data breaches before AI agents process or display potentially compromised results.
• AI data exposure insights provide invaluable visibility into who accessed what information through AI interactions, facilitating rapid investigation. This includes detailed visibility into AI-generated responses, enabling security teams to readily detect and remediate any emerging security risks.

Control, monitor, and protect AI agent data interactions at scale

Sentra’s Data Security for AI Agents is purpose-built to secure AI-powered assistants, such as Microsoft Copilot and OpenAI ChatGPT Enterprise, which connect to vast repositories of corporate knowledge.

Without the implementation of granular access controls and robust data governance, these intelligent agents could inadvertently expose confidential data to employees or users lacking the necessary authorization.

Sentra directly addresses this critical vulnerability by ensuring AI agents meticulously enforce identity-based access rules and monitor for sensitive information leakage, preventing unauthorized disclosures.

Data Security for AI Agents helps to solve many use cases that include the ability to:

• Enforce proper data access controls and prevent data leakage by copilot agents
• Protect and control data used for training or inference by AI models
• Detect model sprawl, shadow AI agents, and govern AI & ML usage for compliance
• Reduce AI inference risks by detecting LLMs that draw and share conclusions with sensitive information

With Sentra’s Data Security for AI Agents solution, large enterprises can confidently embrace the transformative power of AI innovation, secure in the knowledge that their valuable data assets are protected by an purpose-built data security platform.