Product launches

Kong Konnect launches AI Cost Management

Tracks and attributes agent spending across teams, applications, and use cases rather than showing aggregated cloud invoices.

Kong · 4 Nov 2026 · 4 sources · Also reported independently

CrowdStrike launches Agentic Adversary Emulation and mobile threat app

AI-powered red teaming using SafeMind simulates attacks and validates defenses; both now in early access.

CrowdStrike · 4 Nov 2026 · 3 sources · Also reported independently

Aembit adds security for OpenAI Dots personal agents

Aembit now secures OpenAI Dots personal AI agents, expanding coverage of workforce, coding and personal agents.

Aembit · 4 Nov 2026 · 1 source · Vendor sources only

CrowdStrike adds agent insider threat detection to Agentic SOC

New capabilities include risk scoring and multi-agent investigation tools; most launch Q4 2026, some available now.

CrowdStrike · 4 Nov 2026 · 1 source · Vendor sources only

Kong launches Kong Token Vault for AI agents

Exposes SaaS services like GitHub, Figma and Salesforce as MCP servers via corporate SSO with governed OAuth.

Kong · 4 Nov 2026 · 1 source · Vendor sources only

Token Security launches Base44 Compliance API integration

Base44 Enterprise customers can now bring apps, AI agents, integrations and members into Token Security for identity visibility.

Token Security · 4 Nov 2026 · 1 source · Vendor sources only

Cisco launches Foundry platform for LLM security hypothesis verification

Platform uses 'evidence gate' validation with demonstrated applications on Langflow, MLflow, and PyGoat.

CrowdStrike expands Falcon Adversary OverWatch to AI and Google Cloud

Adds 24/7 managed hunting for AI applications and deeper Google Cloud coverage without separate SKUs.

CrowdStrike · 30 Oct 2026 · 1 source · Vendor sources only

CrowdStrike launches Third-Party IOC Management for Falcon Next-Gen SIEM

Feature enables ingestion and management of third-party threat intelligence alongside CrowdStrike's data.

CrowdStrike · 30 Oct 2026 · 1 source · Vendor sources only

Adversa AI launches Coding Agent Security Platform for real-time protection

The platform blocks attack chains on AI coding agents in under 0.2 milliseconds and is already proven in production.

Adversa AI · 29 Oct 2026 · 3 sources · Reported by multiple independent sources

Hush Security launches Agentic Issues for AI agent risk

Maps security issues to OWASP, NIST and MITRE frameworks with severity, owner and credential exposure data.

Hush Security · 29 Oct 2026 · 1 source · Vendor sources only

Check Point launches Maestro Hyperscale Orchestrator 200

400 GbE, up to 4.5 Tbps threat prevention, 4X capacity for AI hyperscale environments.

Check Point · 29 Oct 2026 · 1 source · Vendor sources only

Cisco introduces Project Quarry for standardized agent defense language

The initiative creates standardized detection language for defending against threats targeting autonomous AI agents.

Palo Alto Networks makes Idira agent identity security generally available in Prisma AIRS

The capability connects identity-based permissions with gateway enforcement for AI agents.

Palo Alto Networks · 27 Oct 2026 · 1 source · Vendor sources only

Cisco’s Privileged Infrastructure Access ties admin and AI agent privileges to the task

Cisco launched Privileged Infrastructure Access, granting administrators and AI agents temporary, task-specific privileges for network infrastructure work. The product addresses security risks from excessive standing access on routers and switches.

Palo Alto Networks launches Unit 42 Continuous Frontier AI Defense and Prisma AIRS on OpenAI Marketplace

Products enable machine-speed defense with Frontier Virtual Patching and AI governance, retiring pre-committed OpenAI spend.

Palo Alto Networks · 23 Oct 2026 · 2 sources · Vendor sources only

Knostic launches agents security platform with Shadow AI focus

Platform controls application business flows from user to project, endpoint to browser, and MCP servers to hooks.

Knostic · 22 Oct 2026 · 4 sources · Also reported independently

CrowdStrike adds PHP2Shell protection to Falcon platform

Detection coverage through behavioral monitoring and exposure management tools for CVE-2026-109618 and CVE-2026-109619.

CrowdStrike · 21 Oct 2026 · 2 sources · Also reported independently

Palo Alto Networks launches Unit 42 Managed Agentic Endpoint Security

New offering addresses risks from AI agents accessing sensitive data and executing commands.

Palo Alto Networks · 19 Oct 2026 · 3 sources · Also reported independently

F5 adds agentic AI protections to Distributed Cloud Bot Defense

F5 introduced enhancements combining persistent device intelligence with agentic AI protections to distinguish legitimate AI agents from malicious bots.

F5 · 18 Oct 2026 · 1 source · Vendor sources only

Cloudflare: Aegis: The AI Firewall

Cloudflare · 16 Oct 2026 · 1 source · Also reported independently

Kong launches KSAI agentic engineering platform

The platform evolved from a plugin marketplace into a workflow platform built around GitHub Actions as its runtime engine.

Kong · 16 Oct 2026 · 1 source · Vendor sources only

Airia expands governance module with Two-Stage Intake and Living Disclosures

Two-Stage Intake lets teams register and assess use cases before building; Living Disclosures publishes notices from live governance data.

Airia · 15 Oct 2026 · 1 source · Vendor sources only

Cisco launches DefenseClaw for physical AI security

Extends AI security governance to robots, drones, factory equipment and building systems.

Hush Security launches Hush Entitlements for AI agent access

Adds access context to non-human identities to reveal cloud permissions and reach.

Hush Security · 14 Oct 2026 · 2 sources · Vendor sources only

Token Security launches Agentic Universe educational hub

Free resource offers videos, guides, checklists, and tools for AI agent security created by practitioners.

Token Security · 13 Oct 2026 · 2 sources · Vendor sources only

MintMCP rebrands to Mintive, launches unified AI agent platform

Platform integrates MCP gateway, LLM gateway, security monitoring, and agent management for production AI agents.

MintMCP · 12 Oct 2026 · 2 sources · Also reported independently

Knostic releases Auto Patcher for OpenAnt

Open-source tool generates vulnerability fixes with evidence-based trust reports and self-challenge capabilities.

Knostic · 11 Oct 2026 · 2 sources · Also reported independently

Vijil launches DART automated red-teaming system for AI agents

DART uses adversarial agents for multi-turn testing, integrates with CI/CD, deploys on-prem or air-gapped.

Vijil · 7 Oct 2026 · 4 sources · Reported by multiple independent sources

Aembit launches security support for personal AI agents like Meta Muse

Aembit's MCP Identity Gateway authenticates agents via Okta OIDC with full activity chain visibility.

Aembit · 7 Oct 2026 · 3 sources · Reported by multiple independent sources

Enkrypt AI integrates with ChatGPT Enterprise Compliance API

Integration scans workspace activity for prompt injection, policy violations, PII, and toxic content.

Enkrypt AI · 6 Oct 2026 · 1 source · Vendor sources only

Cloudflare brings paid access to MCP tools — who controls the agent’s spending?

Cloudflare launched a closed beta of its Monetization Gateway, enabling domain owners to charge AI agents for accessing their services. The move introduces payment infrastructure for Model Context Protocol tools, raising questions about how autonomous agents will manage and authorize spending.

Cloudflare · 1 Oct 2026 · 1 source · Also reported independently

Continuous AI security with F5 AI Guardrails and AWS Bedrock

F5 launched AI Guardrails and AI Red Team to provide continuous adaptive security for AI applications running on AWS Bedrock, introducing three enterprise capabilities.

F5 · 1 Oct 2026 · 1 source · Also reported independently

Cisco launches Secure Firewall 200 Series for branch security

The firewall incorporates AI-powered threat detection, encrypted traffic visibility, SD-WAN integration, and simplified management.

Cloudflare AI Wallets Put Agentic Commerce Trust Frameworks to the Test

Cloudflare has launched agent identity verification and payment tools for AI agents, marking progression from experimentation to practical implementation, though fragmented industry standards around trust frameworks persist.

Cloudflare · 27 Sep 2026 · 1 source · Also reported independently

Zscaler launches AI Endpoint Security for zero trust endpoint protection

Extends AI-powered protection to endpoints applying zero trust to AI applications and agents on devices.

Zscaler · 23 Sep 2026 · 2 sources · Vendor sources only

Palo Alto Networks launches Unit 42 Continuous Frontier AI Defense

New cybersecurity offering designed to counter AI-powered threats.

Palo Alto Networks · 22 Sep 2026 · 4 sources · Also reported independently

SentinelOne expands AWS integration for AI governance on Amazon Bedrock

SentinelOne integrates its AI runtime security with Amazon Bedrock AgentCore to monitor AI agents, enforce policies, and detect AI-related threats.

SentinelOne · 16 Sep 2026 · 1 source · Also reported independently

Giskard-oss Giskard-checks/v1.0.4 — 🐢 Open-Source Evaluation & Testing library for LLM Agents

Giskard releases version 1.0.4 of giskard-checks, an open-source library for evaluating and testing large language model agents.

Giskard · 15 Sep 2026 · 1 source · Also reported independently

Aurascape releases September 2026 product update with enhanced AI security

New capabilities include improved sensitive data protection, model-level policy enforcement, and clearer conversation records for monitoring.

Aurascape · 1 Sep 2026 · 1 source · Vendor sources only