Securing autonomous AI agents: agent identity and permissions, guardrails on tool use, runtime protection, and discovering which agents an organization runs.
MarketingProductR&D
AI security posture management
Inventory of AI models, apps and data flows; misconfiguration and risk scoring; AI governance and compliance (EU AI Act, ISO 42001, NIST AI RMF).
ProductMarketing
Funding, acquisitions and partnerships
Funding rounds, acquisitions, strategic partnerships, cloud-marketplace listings, and major customer announcements.
MarketingProduct
MCP and tool-integration security
Model Context Protocol (MCP) servers, agent tool and plugin integrations, and how vendors scan, gateway, or secure them.
ProductR&D
Positioning and category claims
New taglines, category names a vendor is trying to own, comparative claims against competitors, analyst recognition, and pricing or packaging changes.
Marketing
Prompt injection and jailbreak research
New attack techniques, red-teaming results, vulnerability disclosures, and detection approaches for prompt injection and jailbreaks.
Zenity announced support for securing AI agents across Microsoft's ecosystem including GitHub Copilot, Copilot Studio, and other tools, with integration into Microsoft Sentinel.
“Zenity now secures AI agents across the @Microsoft ecosystem. 🛡️
GitHub Copilot, Copilot Studio, Cowork, Claude Enterprise, Foundry, and Agent 365, with high-fidelity alerts in Microsoft Sentinel”Quoted from the original
Marketing
Zenity is positioning itself as the security layer for Microsoft's agentic AI portfolio, potentially claiming category ownership over agent-focused defense in enterprise platforms.
Product
Zenity now covers agents in Copilot Studio, Claude Enterprise, and Foundry; we should evaluate our coverage of these platforms and their agent security model differences.
Zenity published a blog post describing a five-level maturity model (HAMMER) for securing homegrown AI agent development, emphasizing that security must be embedded in design, testing, and release pipelines rather than added late.
“Secure AI agent development is the area where the homegrown, code-first distinction matters most. When an organization buys a SaaS agent, the vendor owns the build, the guardrails, and the release pipeline. When an organization builds its own, every one of those belongs to the organization”Quoted from the original
Product
Zenity is positioning AI agent security maturity as a framework with clear stages, moving from ad-hoc to automated guardrails and policy gates in CI/CD pipelines.
R&D
Zenity highlights threat-modeling for agents, MCP vetting, guardrail standardization, AI-SBOM inventory, and runtime red-teaming as technical security practices, plus protecting configuration from agent self-modification.
Straiker's Head of AI announced benchmark results showing their runtime detection engine catching 98.9% of attacks with a 1.0% false alarm rate against Claude Opus 5.5, GPT-6.1 Sol, and Gemini 3.1 Pro.
“Straiker's runtime detection engine vs. Claude Opus 5.5, GPT-6.1 Sol, and Gemini 3.1 Pro as inline security judges”Quoted from the original
Product
Straiker is positioning runtime detection with strong quantified performance claims that could influence buyer evaluation of detection capabilities.
R&D
Their runtime detection engine uses inline security judges with LLMs; the specific integration approach and attack detection method warrants technical investigation.
HiddenLayer CEO Chris Sestito published commentary on AI security threats, focusing on how bad actors can exploit AI systems and the importance of controls in production AI deployment.
“Every single person with a bad motivation is now a capable hacker”Quoted from the original
Marketing
HiddenLayer is positioning itself as the voice of reason on AI risk, distinguishing between realistic threats (bad actors exploiting uncontrolled AI) and hype (rogue models).
Product
HiddenLayer claims a jailbreak they published remains effective across major models, suggesting models lack sufficient built-in safeguards and reinforcing demand for external control layers.
Vanta has acquired Pillar Security, a startup that discovers, tests, and secures AI agents at runtime; the combined entity will integrate Pillar's agent security capabilities with Vanta's trust and compliance platform.
“Pillar discovers, tests and secures AI applications and agents across their lifecycle, and blocks unauthorized actions at runtime”Quoted from the original
Marketing
Vanta is positioning itself as the 'leading Agentic Trust Platform' and frames agent security as inseparable from broader trust and compliance—a category expansion beyond traditional governance.
Product
Pillar brings runtime agent protection, vulnerability research (CVSS 10 findings), and SAIL framework covering 91 AI risks; integration with Vanta's Trust Graph will connect policy, testing, and enforcement across agent lifecycles.
Lasso Security published a blog post defining rogue AI agents, documenting recent incidents from Anthropic and OpenAI, and positioning their Intent Security Engine as a runtime detection and prevention platform.
“A rogue AI agent is an agent whose actions have moved outside the purpose it was built and authorized for”Quoted from the original
Product
Lasso is narrowing their positioning to agent-specific runtime detection with intent-drift and goal-misalignment monitoring as differentiators against traditional attack-pattern matching.
R&D
Lasso's Intent Security Engine uses behavioral baselines from agent configuration and interaction history, with sub-50ms inline policy enforcement at proxy/gateway layers and 300,000+ OWASP-mapped red-team payloads.
WitnessAI published a blog post explaining how to operationalize the OWASP AI Testing Guide through a five-step lifecycle program and positioning their platform as a complementary runtime enforcement solution.
“WitnessAI provides complementary production controls that can help address three areas beyond the guide's testing methodology: AI activity discovery, runtime data protection, and agentic tool governance”Quoted from the original
Marketing
WitnessAI frames itself as closing OWASP Testing Guide gaps through runtime enforcement, intent-based policies, and agentic security—positioning against test-only approaches.
Product
WitnessAI emphasizes network-level AI discovery, shadow AI detection, MCP/tool governance, data protection guardrails, and bidirectional runtime defense as production-stage capabilities beyond pre-deployment testing.
Zenity Labs disclosed AgentCorruption, a chain of security flaws in Amazon Bedrock AgentCore that allowed attackers to compromise all agents in an AWS account and region through a single prompt to a public-facing agent.
“Using a single prompt to one public-facing agent, researchers took over all AgentCore agents within the same AWS account and region, accessed private conversations and cloud credentials and persistently manipulated agent behavior”Quoted from the original
Product
AgentCore's overprivileged default roles and lack of agent isolation enable lateral movement across agents; enforcing least-privilege execution and agent-level segmentation are critical product differentiators.
R&D
The attack exploited IMDS credential retrieval, overprivileged IAM roles shared across agents, internal agent enumeration APIs, and memory manipulation for persistence; IMDSv2 and reduced role permissions are necessary mitigations.
Straiker published a technical security analysis of a vulnerability chain in Cursor AI affecting indirect prompt injection, sandbox escape via shell builtins, and remote tunnel hijacking that enables persistent remote code execution.
“Cursor's command sandbox is bypassable with a single line. The IDE-level command parser (shouldBlockShellCommand) is blind to shell builtins like export and cd, allowing an attacker to escape the workspace scope and write to arbitrary locations in the user's home directory even with all protections enabled”Quoted from the original
Product
Cursor's sandbox parser blindness to shell builtins and unrestricted home-directory write scope allows attackers to escape workspace isolation and establish persistence through simple repository-based prompt injection.
R&D
The vulnerability combines three technical failures: command parser blind spots on shell builtins (export, cd, source), macOS seatbelt configured to permit home-directory writes, and an unauthenticated tunnel binary with MsgPack RPC spawn method enabling full shell access via hijacked GitHub device-code OAuth flow.
Was this useful?
Product
R&D
Product
R&D
HiddenLayer asserts that guardrails should not originate from model builders, implying architectural separation between model providers and security/control infrastructure is a core design principle.