News for today:
- Closing the AI exposure gap. XM Cyber added three capabilities to its continuous exposure management platform for securing AI – visibility, attack path mapping, and governance and compliance. These are offered as part of an integrated platform, not siloed tools. On attack path mapping – XM Cyber’s Attack Graph Analysis now extends to in-application AI and MCP server exposures, enabling security teams to understand exactly how exposures in AI development and training resources can be chained together to compromise business-critical data. This capability cements XM Cyber’s standing as the only vendor that can validate complete attack paths traversing from internet-facing exposures to cloud AI models to on-premises databases and industrial systems, crossing hybrid environment boundaries that siloed tools cannot see. XM Cyber
- Intent-based AI security for AI agents. Proofpoint released its new Proofpoint AI Security solution, for securing how humans and AI agents use AI. Proofpoint AI Security [works by] applying intent-based detection models, continuously evaluating whether AI behavior, initiated by a human or an autonomous agent, aligns with the original request, defined policies, and intended purpose. By analyzing the full semantic context of AI interactions, the solution flags misaligned or high-risk actions in real time, before damage occurs, such as non-compliant communication or data loss. The new solution builds on Proofpoint’s Acuvity acquisition last month. Proofpoint
- Securing the deployment of AI agents and automated workflows. 1Password announced 1Password Unified Access, for managing credentials, secrets, and machine identities to support the secure deployment of AI agents and automated workflows. Unified Access is built on a clear operating model: discover existing agents and credentials, secure them and continuously authorize access, and audit every action across human and AI agent identities. Unified Access enables security teams to discover, secure, and audit (audit isn’t available immediately). Coming later in 2026: 1Password will expand Unified Access to issue scoped credentials to agent and machine workloads at runtime, further reducing persistent access and strengthening governance as AI-driven automation scales. 1Password
- Low trust in sharing personal data with AI tools. New research from Malwarebytes highlights concerns about what AI tools do with personal data. While AI use is becoming increasingly common, 90% of respondents said they’re worried about AI using their data without consent and 91% support national laws regulating personal data use. The findings highlight growing distrust around AI, with many people scaling back AI use and taking proactive steps to protect their privacy. Malwarebytes
- Discovering, monitoring, and securing the use of unauthorized AI tools. SailPoint launched Shadow AI Remediation, for detecting, monitoring, and securing the use of unauthorized / non-corporate approved / shadow AI tools. The release of Shadow AI Remediation is an important milestone for SailPoint’s comprehensive real-time AI governance and security framework. This framework unifies Agent Identity Security, Machine Identity Security, Data Access Security, and now Shadow AI Remediation to provide a holistic approach to securing AI. By integrating AI tool usage activity into SailPoint’s Identity Security Cloud, organizations can enrich their identity graph, adding important context that enables more intelligent decisions about access and risk. SailPoint
- Posture management for endpoints and identity. Huntress added two new posture management products to its Agentic Security Platform, covering endpoints and identities (in Microsoft 365). The endpoint posture capabilities are brand new; the identity ones are built on the Inside Agent acquisition. Huntress uses threat intelligence from its AI-centric security operations center (SOC) and millions of protected endpoints and identities to continuously define, deploy, and maintain security controls on behalf of customers. By providing the security expertise and technical capabilities that have previously limited access to security posture management tools, the products enable organizations to reduce the number of incidents they face while ensuring they maintain a consistent, hardened security posture across environments. Huntress
- Email security telemetry integration with CrowdStrike Falcon. Check Point introduced a new connector to enable its email security telemetry to flow into the Falcon Next-Gen SIEM from CrowdStrike. The integration allows security teams to ingest and correlate email-based activity with downstream identity and endpoint signals during investigations, helping connect activity from inbox to endpoint within a single workflow. Available immediately. Check Point