News for today:
- [Cayosoft] Launches Microsoft Migration Services. Cayosoft, in partnership with XMS Solutions, introduced a new migration service for Microsoft identity platforms. The Microsoft Migration Service offering provides end-to-end migration services for Active Directory, Microsoft 365, Exchange, SharePoint, Teams, and related identity infrastructure, combining proven execution with pre- and post-migration capabilities. XMS offers migration delivery expertise; Cayosoft the platform. Cayosoft
- [Horizon3.ai] Autonomous AI cyber defense is now safe to deploy. Horizon3.ai extracted stability as an outcome from AI models and embedded that stability in its own platform – in order to make autonomous AI cyber defenses safe to deploy. Horizon3.ai’s research addresses this challenge with a new tool-mediated architecture that makes stability a property of the system itself, not the underlying AI model. The AI remains responsible for strategy, but every action is constrained to a finite, pre-approved catalog and executed through deterministic, validated tools. This ensures the system remains controllable and stable, even under adversarial pressure. The full research report is available. Horizon3.ai
- [Jitterbit] Deep Message Inspection. Jitterbit added MCP capabilities to its Harmony Platform, including measures that bring real-time security across an AI ecosystem. Jitterbit’s Deep Message Inspection (DMI) technology sits across the entire Harmony platform and monitors AI requests, prompts and data exchanges to identify at-risk AI executions, evaluate these against the company’s security policies, then take decision action to mask sensitive data, introduce human oversight or fail the transaction — all in real time. Jitterbit
- [Seclore] Protecting sensitive data in AI systems. Seclore moved its ARMOR AI-DLP solution to GA, for protecting sensitive data across all aspects of AI systems and interactions. AI-DLP identifies sensitive data at the interaction layer in real time between users and applications on one end and AI agents and models on the other, masking it in a bi-directional manner using context-preserving tokenized values before the data reaches any model. The AI can fully reason over the masked data and return results, while real values remain secured within the enterprise environment. On the way back, when the model returns the response, the data is de-tokenized with the original values reinserted to provide complete, useful outputs. Seclore
- [Netskope] Agents for end-to-end workflows. Netskope announced Netskope One AgentSkope, which enables agentic operations to support security and networking teams using the Netskope One Platform. There are six agents available initially (e.g., DLP, insider threat analysis, access analysis) – with more to come. AgentSkope addresses systemic capacity issues, removing complexity and acting as an autonomous force multiplier for both security and networking teams. AgentSkope provides the foundation to build and release agents quickly and efficiently, creates a common set of security, privacy and GRC controls to protect customers uniformly across the platform, provides a consistent experience and tracks and monitors agent utilization. Netskope AI agents help organizations to drastically reduce time spent on manual workflows, from policy creation through to triage, investigation and troubleshooting, allowing organizations to adapt their defenses at the speed of business. Netskope