Cloudbrink launches AI security platform for enterprise

Veraify has launched an AI-native security platform for enterprise AI adoption under a new Cloudbrink brand.

The platform is aimed at companies trying to manage AI tools, agents and assistants across corporate systems while protecting sensitive information. It combines endpoint monitoring, policy enforcement, data protection and secure connectivity in one system.

The launch reflects a broader shift in corporate security as businesses contend with employee use of public generative AI tools, embedded AI features in software and autonomous workflows that can access internal data directly. Older security architectures built around browsers and central inspection points do not provide enough visibility into those interactions, according to Cloudbrink.

Research cited from Gartner found that 69% of organisations suspect or have evidence that employees are using prohibited public generative AI tools. That has increased concern over so-called shadow AI, where workers may enter customer information, employee records, code or proprietary business data into outside services.

Platform focus

Veraify is designed to identify approved and unapproved AI services, agents and copilots used across endpoints. Policy enforcement takes place on the source device, with all AI traffic captured by an agent so it cannot bypass governance controls.

The system also includes data loss prevention functions intended to stop private information, intellectual property and code from leaving the organisation through AI prompts, uploaded files or images. It includes a built-in catalogue for personally identifiable information and can inspect images within files sent to AI systems.

Another part of the product is designed to detect new AI platforms and agents without requiring customers to set up new detection rules manually each time. This is intended to help security teams keep pace as new services appear.

Veraify also applies policy across both human users and AI agents from a single control plane, according to Cloudbrink. That means identity, device and data-based controls can be managed together rather than through separate systems.

Customer view

Cogent Security is among the early users cited by the company.

"As we evaluated AI security platforms, we found that many approached the problem using architectures designed for SASE or web browsers," said Aziz Kapadia, Field CISO, Cogent Security.

"They could inspect browser traffic and some app traffic, but they weren't built to provide visibility into the growing number of AI agents, desktop applications and embedded AI capabilities running directly in apps and on endpoints. We selected Veraify powered by Cloudbrink because its architecture was designed for where enterprise AI security is actually heading, not where it was yesterday," Kapadia said.

Security shift

Cloudbrink is positioning Veraify as a response to a change in how enterprise systems are accessed. Over the past decade, security has largely focused on people logging into applications, while AI agents and assistants now interact directly with data, application programming interfaces and infrastructure, the company argues.

That change has prompted debate in the security market over whether existing zero-trust network access and secure access service edge products can be adapted for AI-heavy workplaces, or whether a different architecture is needed. Cloudbrink's approach builds on its secure connectivity technology and extends it into AI governance and data protection.

"Every major technology shift creates a new security architecture. The PC created endpoint security. The cloud created SASE. AI requires something different," said Prakash Mana, Chief Executive Officer, Cloudbrink.

"AI agents and assistants are becoming digital workers with access to enterprise data, but most companies have very limited visibility or control over what this AI technology is doing. We created Veraify to give enterprises the confidence to adopt AI safely," Mana said.

Built on Cloudbrink

Veraify incorporates and expands the SAFE AI functions that Cloudbrink introduced earlier. The new platform adds broader AI visibility, governance, endpoint-aware protection and secure access in a unified system, according to the company.

Cloudbrink said the product also avoids forcing traffic through centralised inspection points, a design choice intended to preserve user experience while still applying zero-trust access and AI-aware controls. Performance has become a practical issue for corporate AI security tools, as employees may look for ways around them if they introduce delays into common workflows.

"Cloudbrink provides the technology foundation. Veraify defines the next-generation platform for securing AI adoption," said Manohar Reddy, Head of Products, Cloudbrink.

"Veraify extends the foundation we built with Personal SASE into this AI-native era, combining secure access, AI visibility, data protection and performance into one platform. With Veraify, customers do not have to choose between enabling AI innovation and protecting the business. Organisations can get the controls they need to govern AI adoption while preserving the fast, seamless experience users expect," Reddy said.

Veraify is available now.