Your findings list tells you what's wrong. It can't tell you what's reachable.

Attack Paths works from a different input: the actual shape of your estate. Which AWS resources are exposed, which roles can be assumed by whom, which principals can read what. Those relationships are what a path is built from, whether or not anything along the chain has ever been flagged as a problem.

It's live in Infrastructure Discovery now, and there's nothing to install. If your AWS accounts are already connected, it's in your dashboard.

What's new

Paths built from configuration, not from CVEs A path exists because something is exposed, some permission is broader than intended, and something valuable sits at the end of it. Findings from Security Hub, Inspector, and GuardDuty can raise a path's priority, but they never create one and never gate one. Plenty of the paths worth your time have no finding attached anywhere along the chain.

Ranked paths, shown as a map Every path runs from an entry point, through any pivots along the way, to a crown jewel. You get the whole chain as a map, with a severity band and a plain-English reason it landed where it did.

Crown jewels you choose yourself You decide what counts as high value in your estate. Pin and unpin assets from the console, and path scoring reorganizes around your choices instead of our assumptions about your business.

Key nodes: entry points and choke points Entry points are ranked by what they actually open up. Choke points are the nodes that the most paths run through, which is usually where a single change does the most work.

Identity, not just network reachability Paths follow IAM as closely as they follow the network: role chaining, pass-role, policy attachment, self-escalation, and principals holding effective admin through AWS managed policies.

Why it matters

A severity list ranks problems one at a time, and it's genuinely good at that. What it can't do is tell you that a role you granted a vendor two years ago, with "read-only" in its name and a wildcard in its policy, sits one hop from every secret you own. There's no CVE on that role and nothing about it is broken. Someone made a reasonable call and the blast radius grew around it quietly. That's a path, and paths are what this is for.

Where to find it

Infrastructure Discovery → Attack Paths. 

Attack Paths is included in your Infrastructure Discovery license. No separate SKU, no add-on charge.

Full details in the Infrastructure Discovery documentation.

Wallarm is pleased to announce our new Attacks interface.

There’s nothing to install. The new interface will be automatically added to your existing dashboard.

Designed to match the way security teams investigate threats, the new experience makes it easier to identify attacks, spot patterns, and drill into the details that matter.

What's new

Interactive Dashboard
Navigate attack trends and spot anomalies across your traffic. Apply filters directly from charts, tables, and statistics to quickly narrow the scope of your investigation.

Customizable Views
Save views with your preferred columns, filters, time range, and grouping. Create dedicated dashboards for specific attack types, applications, hosts, or incidents.

Attack Investigation Flow
Drill into any attack to inspect request-level details, including source IP, host, URI, payload, parameters, classifications, headers, and the full raw request.

Interactive Request Grouping
Group attacking requests by attack type, source IP, host, path, or a custom combination to reveal patterns that are difficult to spot one request at a time.

Customizable Export
Export attack data using your selected filters, columns, time range, and grouping. Tailor exports for incident response, reporting, compliance, or offline analysis.

Your API security is already covered. What's changed is what's behind those APIs. AI agents are making decisions, accessing data, and calling external services right now, and most security teams can't see any of it. Not because they're not paying attention, but because no tool was built to show them.

Today that changes. We're launching the Wallarm AI Control Platform: two products that close the loop from estate-wide discovery through runtime enforcement through continuous compliance evidence. No new vendor to onboard. It extends the platform you already run.

What's shipping today

Infrastructure Discovery

Connect your AWS accounts once and get a continuously updated inventory of everything in your estate. EC2, VPC topology, EKS clusters, Lambda functions, API Gateway, IAM, and Bedrock models and agents, across every account and every region, in one searchable table.

• Cross-account discovery via IAM role assumption, no write permissions required
• Live relationship graph with blast-radius traversal and attack-path analysis
• AWS Security Hub findings sync, placed on the graph node they affect with full asset context
• Field-level drift detection between every scan, with CloudTrail creator attribution on every asset
• Customer-authored detection and triage rules in Common Expression Language
• Scheduled and on-demand scans; policy audit log for every triage decision

AI Hypervisor

A Kubernetes DaemonSet that instruments every AI workload at runtime via a mutating admission webhook, with zero application code changes. Label a namespace, and coverage begins within minutes. Works across Python, Go, Node.js, Java, Ruby, and generic containers.

• Parses every major model provider: Anthropic, OpenAI, AWS Bedrock, Azure OpenAI, Google Gemini, and more
• Attributes every LLM call back to the originating user or session, across internal service hops
• Real-time sensitive data detection: credit cards, SSNs, passport numbers, API keys, JWT tokens
• Session kill switch by user subject or W3C trace ID, enforced at the kernel, no restart required
• Agent behavior certificates that pin and enforce what each agent is permitted to do
• Continuous compliance report: AI inventory, coverage heatmap, session logs, PII egress records
• SIEM, SOAR, and ticketing integrations for findings and policy violations

EU AI Act enforcement starts in August 2026. If that's on your radar, AI Hypervisor generates the compliance evidence you'll need continuously, not on demand when an audit appears. Getting it running now means you won't be assembling spreadsheets in July.

Learn more in our documentation for AI Hypervisor and Infrastructure Discovery, or request a demo. 

MCP servers provide structured access to privileged systems for AI agents and external integrations. That’s great for enabling things like automated project management, embedded billing, and AI-driven developer workflows. But it also creates a new attack surface that existing security tools were not designed to secure. 

That's why we built MCP Protection, a feature that extends the discovery, visibility, and control you already rely on for APIs to your externally exposed MCP servers.

With MCP Protection, you can:

• Discover exposed MCP servers, tools, and resources
• Correlate MCP and user sessions for full activity visibility
• Enforce granular access and schema-based controls for tool usage

MCP Protection is included in Wallarm Advanced API Security.

Check out this overview of AI and MCP capabilities in the doc hub for more info.

API Discovery now detects how every endpoint in your inventory is authenticated, and how consistently that authentication shows up in real traffic. Bearer tokens, API keys, AWS Signature v4, Basic, cookie-based auth, and the rest. It's all classified, tracked, and filterable.

If you've ever wondered whether that one internal endpoint really requires a token or just usually does, this is for you.

What's new

Per-endpoint authentication status. Every endpoint gets classified as Consistent, Partial, or Missing, based on a rolling 7-day window of production traffic. No guessing, no spec-reading. Just what's actually happening on the wire.

A filter for unauthenticated endpoints. Open the inventory, flip the filter, and you've got a list of exposed APIs in seconds. This is the one most teams will want bookmarked.

Per-parameter coverage on the endpoint detail page. You can see exactly which header, cookie, or parameter is carrying the credential, and what fraction of requests include it. Useful when "Partial" shows up and you need to know why.

Why it matters

Missing auth is one of those problems that's obvious in hindsight and invisible until something goes wrong. Specs say one thing, traffic says another, and the gap is where breaches live. This grounds the answer in what your APIs are doing right now, not what a YAML file claims they do.

Availability

Upgrade your filtering nodes to NGINX Node 6.10.0 or Native Node 0.23.0 (or newer) to start receiving authentication data.

Full guide: docs.wallarm.com/api-discovery/authentication

Sometimes you need visibility into your APIs without touching production traffic.

Wallarm’s TCP traffic analysis engine now supports AWS VPC Traffic Mirroring, so you can mirror traffic from your AWS environment to a Wallarm Node and see what’s happening across your APIs, without adding latency or changing your request path.

What you get

Mirror traffic directly from your AWS Elastic Network Interfaces (ENIs) and unlock:

• API Discovery — Know what APIs you actually have
• API Session analysis — Understand how they’re being used
• Attack detection — See threats targeting your APIs

All out-of-band. No impact to live traffic.

What’s new

• Support for Geneve-encapsulated traffic (used by AWS Traffic Mirroring)
• Support for VXLAN encapsulation
• Native compatibility with AWS traffic mirroring workflows

Why it matters

Some environments require absolutely zero added latency. Others just prefer to observe instead of actively block.

This gives you a clean way to get API visibility and threat detection in AWS without changing how your traffic flows.

Get started

Check out the documentation to start mirroring your AWS traffic to Wallarm.

The new Wallarm NGINX Ingress Controller 7.0 is now available. This release is based on the open-source NGINX Ingress Controller and is fully integrated Wallarm security services. It replaces all previous versions based on the Community NGINX Ingress Controller, which is end of life in March 2026.

Who should upgrade: Teams currently running the Community-based Wallarm Ingress Controller are strongly encouraged to migrate. The upstream Community NGINX project has been retired. If you are running previous versions of the Wallarm NGINX Ingress Controller, there will be no further security updates, bug fixes, or updates of any kind. 

A migration guide from Wallarm is available.

Full deployment guide: https://docs.wallarm.com/7.x/admin-en/installation-kubernetes-en/

Your attack surface doesn’t stop at domains. Now your visibility doesn’t either.

We’ve rolled out a set of API Attack Surface Management (AASM) updates that expand what you can scan, speed up how you validate fixes, and make it easier to share what actually matters.

What’s new?

• Scan IPs and network ranges
  Go beyond domains and hosts. You can now scan single IPs, subnets, and CIDR ranges to uncover more internet-facing assets and hidden exposure.
• Security Report Card (PDF)
  Get a clean, executive-ready snapshot of your security posture—plus your 10 most critical findings—in a concise, shareable format.
• Rescan a single host
  No need to rerun everything. Quickly rescan one asset to validate fixes and confirm remediation.

Why it matters
Attack surfaces are messy. Assets live outside neat domain boundaries, fixes need verification, and stakeholders want clear answers.

These updates help you:

• Find more assets you didn’t know you had
• Validate fixes faster without slowing down your workflow
• Communicate risk clearly to both technical and non-technical audiences

Where to find it

• IP and network scans → AASM Configuration
• Security Report Card → Share Report widget
• Host rescan → Available per asset

Read more in our documentation.

You shouldn’t have to guess who changed what in your Wallarm configuration. Now you don’t.

We’ve introduced a fully redesigned Activity Log that gives you a clear, attributable record of every action across your Wallarm platform so you can move faster, stay compliant, and actually trust what you’re seeing.

What’s new?

• Full event coverage: Track configuration changes, access updates, and system activity in one place
• Clear attribution: See exactly who performed each action and when. No more detective work
• Operational clarity: Quickly understand what changed and follow up with the right person
• Compliance-ready logging: Built to support audit requirements across regulated industries
• Enterprise-friendly: Designed for teams with multiple users and shared responsibility

Where to find it Settings → Activity Log

Documentation

Better inputs. Better outputs. Better control.

We've improved Wallarm's Schema-Based Testing with updates that make it easier to plug into your existing workflows and test APIs the way you actually run them.

What’s New

RAML Support
Upload RAML files as input. We’ll automatically convert them to OpenAPI specifications so you can get testing without rework.

JUnit Output for CI/CD
Export test results in JUnit format for clean CI/CD integration. Your pipelines stay happy. Your developers stay informed.

Full Request & Response Log Export
Need visibility? Export complete logs for all requests and responses to support troubleshooting, validation, and reporting.

mTLS Support
Mutual TLS authentication is now supported, so you can test APIs that require stronger identity verification.

Because the fastest way to reduce API risk is to test it before someone else does. You can learn more about these capabilities in the documentation.