Impart Resources
Impart Product Update - Nov 2025
We’ve delivered a major round of upgrades across the Impart platform, introducing new AI Bot/MCP and LLM Protection dashboards, a refreshed and more intuitive App Experience, a high-performance Inspector v0.42.0 release, expanded Inspector Metrics for deeper operational visibility, and new SQLi and XSS version control, allowing teams to choose between Detection Version 1, Version 2, or always use the latest release. These updates make it easier than ever to understand AI-driven traffic, configure protections with clarity, manage detection behavior with precision, monitor system performance, and optimize your entire Impart deployment.
AI Agent Security vs. AI Agent Protection
The AI security market is moving fast to respond to predictions that over 40% of enterprise applications will have embedded AI agents by the end of 2026. Every vendor is packaging their answer under the label "AI agent security." The problem is that "AI agent security" means something different depending on who's selling it. This post defines what each layer does, where each one runs out of road, and why the architecture needs both.
Filters
Smarter, Clearer Rules: Now with Condition Groups, Reordering, Multi-line Descriptions & Numeric Operators
We’ve made it easier to build and manage rules with new features designed for clarity and control. You can now group conditions for more powerful logic, reorder them with ease, add multi-line descriptions for better documentation, and use numeric operators in CTX-based rules to support advanced comparisons.
Understanding EchoLeak: What This Vulnerability Teaches Us About Application Security
The recent disclosure of EchoLeak by Aim Labs marks a significant milestone in AI security research. As the first documented zero-click exploit targeting a production AI system, it offers valuable insights into the emerging threat landscape that security professionals need to understand and prepare for.
.avif)
.avif)
Introducing the runtime rule library
Instead of spending time writing and testing custom logic, detection and response teams can now tap into a curated set of rules tailored for modern cloud applications. These aren't your standard WAF rules — they're full-featured, deeply customizable, and built to align with your business logic. Even better, the rule library has rules of different types - including rule recipes (rule builder based) or rule scripts (code based)!
Custom Client Identifiers: Smarter Client Tracking for Security Teams
Relying on IP addresses alone isn't enough to reliably track or distinguish clients—especially in modern, complex environments. That’s why we’ve introduced Custom Client Identifiers, giving your team the flexibility to define what a “client” really means based on your specific use case.Whether it’s IP + User-Agent, IP + API Key, or Token + Product Identifier, you can now tailor client identification to suit your architecture and security goals.
.avif)
.png)
.avif)
