Sentra launches Shadow AI DLP to block sensitive data

Sentra has launched Shadow AI DLP within its AI Data Readiness Platform, aiming to stop sensitive information from being shared with generative AI applications.

The product is designed to give security and governance teams visibility into employees' use of tools such as ChatGPT, Claude, Gemini and Copilot, while applying controls before data leaves the browser. It combines Shadow AI Discovery and AI Browser DLP within the same platform.

Businesses are under pressure to manage the rapid spread of browser-based AI tools in day-to-day work. Employees increasingly use these services to draft text, write code, summarise documents and analyse internal information, creating a risk that customer records, financial data, source code and regulated information could be pasted or uploaded outside established controls.

That shift has exposed a gap in older security products. Traditional data loss prevention systems, secure web gateways and cloud access security broker tools were built mainly to monitor email, file transfers, network traffic and recognised software-as-a-service applications, rather than direct browser interactions with AI tools.

Sentra says organisations can use its approach to discover which AI applications employees are using and apply policies based on both the application and the classification of the data involved. Those controls can block, allow, log, require justification or prompt for verification when a user attempts to share information with an AI service.

The product also creates an audit trail covering the user, the application, the data classification and the enforcement action taken. Sentra says the system classifies content locally in the browser and does not send original content back to the company or route browser traffic through a new proxy.

Growing concern

The launch comes as boardrooms and security teams pay closer attention to the role of data governance in AI deployments. Sentra pointed to Gartner research suggesting that more than 60% of AI projects will fail to meet business service-level agreements without an AI-ready data practice.

It also cited McKinsey research indicating that organisations seen as strong AI performers identify data, rather than models, talent or technology, as the biggest barrier to scaling AI. Against that backdrop, companies are increasing spending on systems meant to govern how data is prepared, classified and controlled when used with AI tools.

Sentra estimated the AI data readiness market at USD $2.1 billion in 2026 and said it expects that figure to reach USD $15.5 billion by 2030. It argued that employee use of generative AI applications represents one of the most immediate areas of exposure because it often happens outside the visibility of existing controls.

Product strategy

Rather than presenting Shadow AI DLP as a replacement for existing security systems, Sentra positioned it as an extension of current data loss prevention investments. Its wider strategy is to supply data intelligence to other DLP products, using continuously updated classifications and business context to improve enforcement decisions and reduce false positives.

In that model, Shadow AI DLP addresses what Sentra sees as a specific weak point: browser-based interactions with AI applications. This allows organisations using older DLP tools to add controls over AI use without replacing broader systems already deployed across email, endpoints or cloud services.

Yair Cohen, Co-founder and Chief Product Officer at Sentra, said companies are moving beyond deciding which AI tools employees may use to understanding what information they are sharing.

"As AI becomes embedded in everyday work, organizations need to govern not just which AI applications employees use, but what data they're sharing with them," said Cohen, Co-founder and Chief Product Officer at Sentra.

He said older security products make decisions based only on what they can observe when enforcement is triggered, while Sentra is trying to add broader business context from its data platform.

"Traditional DLP solutions make decisions based on what they see at the point of enforcement. Shadow AI DLP extends the intelligence of our AI Data Readiness Platform into the browser, applying business context to every policy decision. The result is more accurate detection, fewer false positives and the confidence to adopt AI without putting sensitive data at risk," Cohen said.