Innovation Hub

Featured Posts

Insights
xx
min read

Introducing Workflow-Aligned Modules in the HiddenLayer AI Security Platform

Insights
xx
min read

Inside HiddenLayer’s Research Team: The Experts Securing the Future of AI

Insights
xx
min read

Why Traditional Cybersecurity Won’t “Fix” AI

Get all our Latest Research & Insights

Explore our glossary to get clear, practical definitions of the terms shaping AI security, governance, and risk management.

Research

Research
xx
min read

Agentic ShadowLogic

Research
xx
min read

MCP and the Shift to AI Systems

Research
xx
min read

The Lethal Trifecta and How to Defend Against It

Research
xx
min read

EchoGram: The Hidden Vulnerability Undermining AI Guardrails

Videos

Report and Guides

Report and Guide
xx
min read

Securing AI: The Technology Playbook

Report and Guide
xx
min read

Securing AI: The Financial Services Playbook

Report and Guide
xx
min read

AI Threat Landscape Report 2025

HiddenLayer AI Security Research Advisory

CVE-2025-62354
XX
min read

Allowlist Bypass in Run Terminal Tool Allows Arbitrary Code Execution During Autorun Mode

When in autorun mode with the secure ‘Follow Allowlist’ setting, Cursor checks commands sent to run in the terminal by the agent to see if a command has been specifically allowed. The function that checks the command has a bypass to its logic, allowing an attacker to craft a command that will execute non-whitelisted commands.

SAI-ADV-2025-012
XX
min read

Data Exfiltration from Tool-Assisted Setup

Windsurf’s automated tools can execute instructions contained within project files without asking for user permission. This means an attacker can hide instructions within a project file to read and extract sensitive data from project files (such as a .env file) and insert it into web requests for the purposes of exfiltration.

CVE-2025-62353
XX
min read

Path Traversal in File Tools Allowing Arbitrary Filesystem Access

A path traversal vulnerability exists within Windsurf’s codebase_search and write_to_file tools. These tools do not properly validate input paths, enabling access to files outside the intended project directory, which can provide attackers a way to read from and write to arbitrary locations on the target user’s filesystem.

CVE-2025-62356
XX
min read

Symlink Bypass in File System MCP Server Leading to Arbitrary Filesystem Read

A symlink bypass vulnerability exists inside of the built-in File System MCP server, allowing any file on the filesystem to be read by the model. The code that validates allowed paths can be found in the file: ai/codium/mcp/ideTools/FileSystem.java, but this validation can be bypassed if a symbolic link exists within the project.

In the News

News
XX
min read
HiddenLayer Selected as Awardee on $151B Missile Defense Agency SHIELD IDIQ Supporting the Golden Dome Initiative

Underpinning HiddenLayer’s unique solution for the DoD and USIC is HiddenLayer’s Airgapped AI Security Platform, the first solution designed to protect AI models and development processes in fully classified, disconnected environments. Deployed locally within customer-controlled environments, the platform supports strict US Federal security requirements while delivering enterprise-ready detection, scanning, and response capabilities essential for national security missions.

News
XX
min read
HiddenLayer Announces AWS GenAI Integrations, AI Attack Simulation Launch, and Platform Enhancements to Secure Bedrock and AgentCore Deployments

As organizations rapidly adopt generative AI, they face increasing risks of prompt injection, data leakage, and model misuse. HiddenLayer’s security technology, built on AWS, helps enterprises address these risks while maintaining speed and innovation.

News
XX
min read
HiddenLayer Joins Databricks’ Data Intelligence Platform for Cybersecurity

On September 30, Databricks officially launched its <a href="https://www.databricks.com/blog/transforming-cybersecurity-data-intelligence?utm_source=linkedin&amp;utm_medium=organic-social">Data Intelligence Platform for Cybersecurity</a>, marking a significant step in unifying data, AI, and security under one roof. At HiddenLayer, we’re proud to be part of this new data intelligence platform, as it represents a significant milestone in the industry's direction.

Insights
xx
min read

Life at HiddenLayer: Where Bold Thinkers Secure the Future of AI

At HiddenLayer, we’re not just watching AI change the world—we’re building the safeguards that make it safer. As a remote-first company focused on securing machine learning systems, we’re operating at the edge of what’s possible in tech and security. That’s exciting. It’s also a serious responsibility. And we’ve built a team that shows up every day ready to meet that challenge.

Insights
xx
min read

Integrating HiddenLayer’s Model Scanner with Databricks Unity Catalog

As machine learning becomes more embedded in enterprise workflows, model security is no longer optional. From training to deployment, organizations need a streamlined way to detect and respond to threats that might lurk inside their models. The integration between HiddenLayer’s Model Scanner and Databricks Unity Catalog provides an automated, frictionless way to monitor models for vulnerabilities as soon as they are registered. This approach ensures continuous protection without slowing down your teams.

Insights
xx
min read

Behind the Build: HiddenLayer’s Hackathon

At HiddenLayer, innovation isn’t a buzzword; it’s a habit. One way we nurture that mindset is through our internal hackathon: a time-boxed, creativity-fueled event where employees step away from their day-to-day roles to experiment, collaborate, and solve real problems. Whether it’s optimizing a workflow or prototyping a tool that could transform AI security, the hackathon is our space for bold ideas.

Insights
xx
min read

The AI Security Playbook

As AI rapidly transforms business operations across industries, it brings unprecedented security vulnerabilities that existing tools simply weren’t designed to address. This article reveals the hidden dangers lurking within AI systems, where attackers leverage runtime vulnerabilities to exploit model weaknesses, and introduces a comprehensive security framework that protects the entire AI lifecycle. Through the real-world journey of Maya, a data scientist, and Raj, a security lead, readers will discover how HiddenLayer’s platform seamlessly integrates robust security measures from development to deployment without disrupting innovation. In a landscape where keeping pace with adversarial AI techniques is nearly impossible for most organizations, this blueprint for end-to-end protection offers a crucial advantage before the inevitable headlines of major AI breaches begin to emerge.

Insights
xx
min read

Governing Agentic AI

Artificial intelligence is evolving rapidly. We’re moving from prompt-based systems to more autonomous, goal-driven technologies known as agentic AI. These systems can take independent actions, collaborate with other agents, and interact with external systems—all with limited human input. This shift introduces serious questions about governance, oversight, and security.

Insights
xx
min read

AI Policy in the U.S.

Artificial intelligence (AI) has rapidly evolved from a cutting-edge technology into a foundational layer of modern digital infrastructure. Its influence is reshaping industries, redefining public services, and creating new vectors of economic and national competitiveness. In this environment, we need to change the narrative of “how to strike a balance between regulation and innovation” to “how to maximize performance across all dimensions of AI development”.

Insights
xx
min read

RSAC 2025 Takeaways

RSA Conference 2025 may be over, but conversations are still echoing about what’s possible with AI and what’s at risk. This year’s theme, “Many Voices. One Community,” reflected the growing understanding that AI security isn’t a challenge one company or sector can solve alone. It takes shared responsibility, diverse perspectives, and purposeful collaboration.

Insights
xx
min read

Universal Bypass Discovery: Why AI Systems Everywhere Are at Risk

HiddenLayer researchers have developed the first single, universal prompt injection technique, post-instruction hierarchy, that successfully bypasses safety guardrails across nearly all major frontier AI models. This includes models from OpenAI (GPT-4o, GPT-4o-mini, and even the newly announced GPT-4.1), Google (Gemini 1.5, 2.0, and 2.5), Microsoft (Copilot), Anthropic (Claude 3.7 and 3.5), Meta (Llama 3 and 4 families), DeepSeek (V3, R1), Qwen (2.5 72B), and Mixtral (8x22B).

Insights
xx
min read

How To Secure Agentic AI

Artificial Intelligence is entering a new chapter defined not just by generating content but by taking independent, goal-driven action. This evolution is called agentic AI. These systems don’t simply respond to prompts; they reason, make decisions, contact tools, and carry out tasks across systems, all with limited human oversight. In short, they are the architects of their own workflows.

Insights
xx
min read

What’s New in AI

The past year brought significant advancements in AI across multiple domains, including multimodal models, retrieval-augmented generation (RAG), humanoid robotics, and agentic AI.

Insights
xx
min read

Securing Agentic AI: A Beginner's Guide

The rise of generative AI has unlocked new possibilities across industries, and among the most promising developments is the emergence of agentic AI. Unlike traditional AI systems that respond to isolated prompts, agentic AI systems can plan, reason, and take autonomous action to achieve complex goals.

Insights
xx
min read

AI Red Teaming Best Practices

Organizations deploying AI must ensure resilience against adversarial attacks before models go live. This blog covers best practices for <a href="https://hiddenlayer.com/innovation-hub/a-guide-to-ai-red-teaming/">AI red teaming, drawing on industry frameworks and insights from real-world engagements by HiddenLayer’s Professional Services team.

research
xx
min read

Agentic ShadowLogic

research
xx
min read

MCP and the Shift to AI Systems

research
xx
min read

The Lethal Trifecta and How to Defend Against It

research
xx
min read

EchoGram: The Hidden Vulnerability Undermining AI Guardrails

research
xx
min read

Same Model, Different Hat

research
xx
min read

The Expanding AI Cyber Risk Landscape

research
xx
min read

The First AI-Powered Cyber Attack

research
xx
min read

Prompts Gone Viral: Practical Code Assistant AI Viruses

research
xx
min read

Persistent Backdoors

research
xx
min read

Visual Input based Steering for Output Redirection (VISOR)

research
xx
min read

How Hidden Prompt Injections Can Hijack AI Code Assistants Like Cursor

research
xx
min read

Introducing a Taxonomy of Adversarial Prompt Engineering

Report and Guide
xx
min read

Securing AI: The Technology Playbook

Report and Guide
xx
min read

Securing AI: The Financial Services Playbook

Report and Guide
xx
min read

AI Threat Landscape Report 2025

Report and Guide
xx
min read

HiddenLayer Named a Cool Vendor in AI Security

Report and Guide
xx
min read

A Step-By-Step Guide for CISOS

Report and Guide
xx
min read

AI Threat landscape Report 2024

Report and Guide
xx
min read

HiddenLayer and Intel eBook

Report and Guide
xx
min read

Forrester Opportunity Snapshot

news
xx
min read

HiddenLayer Creates a Threat Intelligence Team Focused on Thwarting ML Attacks

news
xx
min read

These 5 Austin Tech Companies Raised a Combined $507M in July

news
xx
min read

Machine learning creates a new attack surface requiring specialized defenses

news
xx
min read

Spruce Picked up $26M, Tecovas’ New CEO, and More Austin Tech News

news
xx
min read

Why AI is the key to cutting-edge cybersecurity

news
xx
min read

Managed Security Services Provider (MSSP) Market News: 20 July 2022

news
xx
min read

HiddenLayer Launches Security Solution to Protect AI-Powered Products

news
xx
min read

HiddenLayer emerges from stealth to protect AI models from attacks

news
xx
min read

HiddenLayer Launches the First Security Solution to Protect AI-Powered Products

news
xx
min read

$422.37+ Billion Global Artificial Intelligence (AI) Market Size Likely to Grow at 39.4% CAGR During 2022-2028 | Industry

SAI Security Advisory

Allowlist Bypass in Run Terminal Tool Allows Arbitrary Code Execution During Autorun Mode

When in autorun mode with the secure ‘Follow Allowlist’ setting, Cursor checks commands sent to run in the terminal by the agent to see if a command has been specifically allowed. The function that checks the command has a bypass to its logic, allowing an attacker to craft a command that will execute non-whitelisted commands.

SAI Security Advisory

Data Exfiltration from Tool-Assisted Setup

Windsurf’s automated tools can execute instructions contained within project files without asking for user permission. This means an attacker can hide instructions within a project file to read and extract sensitive data from project files (such as a .env file) and insert it into web requests for the purposes of exfiltration.

SAI Security Advisory

Path Traversal in File Tools Allowing Arbitrary Filesystem Access

A path traversal vulnerability exists within Windsurf’s codebase_search and write_to_file tools. These tools do not properly validate input paths, enabling access to files outside the intended project directory, which can provide attackers a way to read from and write to arbitrary locations on the target user’s filesystem.

SAI Security Advisory

Symlink Bypass in File System MCP Server Leading to Arbitrary Filesystem Read

A symlink bypass vulnerability exists inside of the built-in File System MCP server, allowing any file on the filesystem to be read by the model. The code that validates allowed paths can be found in the file: ai/codium/mcp/ideTools/FileSystem.java, but this validation can be bypassed if a symbolic link exists within the project.

SAI Security Advisory

Data Exfiltration through Web Search Tool

The Web Search functionality within the Qodo Gen JetBrains plugin is set up as a built-in MCP server through ai/codium/CustomAgentKt.java. It does not ask user permission when called, meaning that an attacker can enumerate code project files on a victim’s machine and call the Web Search tool to exfiltrate their contents via a request to an external server.

SAI Security Advisory

Unsafe deserialization function leads to code execution when loading a Keras model

An arbitrary code execution vulnerability exists in the TorchModuleWrapper class due to its usage of torch.load() within the from_config method. The method deserializes model data with the weights_only parameter set to False, which causes Torch to fall back on Python’s pickle module for deserialization. Since pickle is known to be unsafe and capable of executing arbitrary code during the deserialization process, a maliciously crafted model file could allow an attacker to execute arbitrary commands.

SAI Security Advisory

How Hidden Prompt Injections Can Hijack AI Code Assistants Like Cursor

When in autorun mode, Cursor checks commands against those that have been specifically blocked or allowed. The function that performs this check has a bypass in its logic that can be exploited by an attacker to craft a command that will be executed regardless of whether or not it is on the block-list or allow-list.

SAI Security Advisory

Exposure of sensitive Information allows account takeover

By default, BackendAI’s agent will write to /home/config/ when starting an interactive session. These files are readable by the default user. However, they contain sensitive information such as the user’s mail, access key, and session settings. A threat actor accessing that file can perform operations on behalf of the user, potentially granting the threat actor super administrator privileges.

SAI Security Advisory

Improper access control arbitrary allows account creation

By default, BackendAI doesn’t enable account creation. However, an exposed endpoint allows anyone to sign up with a user-privileged account. This flaw allows threat actors to initiate their own unauthorized session and exploit the resources—to install cryptominers, use the session as a malware distribution endpoint—or to access exposed data through user-accessible storages.

SAI Security Advisory

Missing Authorization for Interactive Sessions

BackendAI interactive sessions do not verify whether a user is authorized and doesn’t have authentication. These missing verifications allow attackers to take over the sessions and access the data (models, code, etc.), alter the data or results, and stop the user from accessing their session.

SAI Security Advisory

Unsafe Deserialization in DeepSpeed utility function when loading the model file

The convert_zero_checkpoint_to_fp32_state_dict utility function contains an unsafe torch.load which will execute arbitrary code on a user’s system when loading a maliciously crafted file.

SAI Security Advisory

keras.models.load_model when scanning .pb files leads to arbitrary code execution

A vulnerability exists inside the unsafe_check_pb function within the watchtower/src/utils/model_inspector_util.py file. This function runs keras.models.load_model on a .pb file that the user wants to scan for malicious payloads. A maliciously crafted .pb file will execute its payload when run with keras.models.load_model, allowing for a user’s device to be compromised when scanning a downloaded file.

Stay Ahead of AI Security Risks

Get research-driven insights, emerging threat analysis, and practical guidance on securing AI systems—delivered to your inbox.

By submitting this form, you agree to HiddenLayer's Terms of Use and acknowledge our Privacy Statement.

Thanks for your message!

We will reach back to you as soon as possible.

Oops! Something went wrong while submitting the form.