News for today:
- [M-Files] M-Files for Manufacturing. M-Files launched a new document management system for engineering-intensive manufacturers that both leverages AI and lays the foundation for more effective use of AI capabilities. M-Files for Manufacturing [provides] governed, contextualized information so every drawing, work instruction, and specification is connected to the part, process and asset it supports, creating the AI-ready foundation that improves findability and strengthens trust. M-Files
- [Menlo Security] Google SecOps integration. Menlo Security integrated threat signals from its browser with Google Security Operations, enabling correlation with other threat signals within Google SecOps and the return of a proposed response to Menlo for action. Menlo HEAT Shield Agent inspects live page content with Google’s Gemini models and blocks zero-day phishing at the point of click, ahead of any reputation feed. Each detection can now stream into Google Security Operations , where a customer’s playbook can correlate it against existing telemetry and propose a response. Once a SecOps analyst approves it, Google Security Operations returns the instruction to Menlo’s control plane, which carries out whatever the customer has built it to do, whether that’s a policy change, a blocklist update, or a session containment, or any other action supported by Menlo API. Menlo Security
- [Reco] $55 million from AT&T Ventures (and others). Reco announced an additional $55 million in funding from various investors; total funding is now $140 million. Reco will use the funds to expand sales, partnerships, channels and customer support as large enterprises move from early agent security projects to broader deployment. Reco’s integrations with hundreds of leading AI providers and enterprise applications, including OpenAI, Anthropic, Microsoft Copilot, Salesforce, ServiceNow and Workday, give customers the coverage needed to govern agents where they already operate. Reco
- [Radware] Trusting AI agents. Radware introduced the Radware Agent Trust Management solution to control how AI agents interact with applications; uses intent, behavioral analysis, baselining, etc. Identifying an AI agent alone does not establish whether it can be trusted. Even a verified agent can have malicious intent, deviate from its stated purpose or attempt harmful actions. Radware Agent Trust Management is designed to go beyond relying solely on identity and observed behavior by also directly capturing an agent’s declared intent through active conversation and continuously evaluating whether its actions remain consistent with that intent. This helps organizations distinguish legitimate agent interactions from potentially harmful activity. Radware
- [LastPass] More Max. LastPass expanded its Business Max offering with new visibility and governance capabilities for AI tools, SaaS apps, and websites. AI Monitoring & Protect closes the gap between the speed of AI activity and IT’s ability to see and govern its use. Ninety-two percent of organizations say AI is already in use across their company, yet nearly two-thirds either have significant ungoverned AI use or lack the visibility to identify it, posing serious security concerns. With its expanded Business Max capabilities, LastPass administrators are empowered to govern apps and sites by category, help block visits to malicious sites, and identify and warn users before sensitive data is shared with AI tools. LastPass
- [Cribl] Controlling AI costs. Cribl launched StreamAI, an enterprise AI gateway with a model router; the gateway routes AI prompts and workloads to the best model (characteristics, pricing). As more enterprises ramp up AI use in daily operations, spending on AI inference is becoming harder to predict and control. Infrastructure costs can spike with little warning, particularly when agents are coordinating multi-step tasks with minimal human intervention. StreamAI delivers real-time cost control by providing reliable, high-performance inference across proprietary and open-source models while enforcing budgets and spending cutoffs. Cribl
- [Delinea] AI tools going rogue. New research from Delinea finds systemic violation of sensitive data access policies by AI tools, as well as intentional AI policy violations by employees. 87% of IT leaders say an AI tool or agent accessed sensitive data beyond its intended scope in the past year. The report also found that fewer than one in five can detect a scope violation as it happens …. Employees are already routing around [AI policies]: 76% say they have bypassed the required approval process for using AI at some point. Delinea