#AI Security
35 articles with this tag

Cloudflare's AI Security Blueprint
Cloudflare details its model-agnostic AI security harness architecture for scalable vulnerability discovery and validation.
Databricks Unveils Open AI Governance
Databricks is expanding its Unity AI Gateway to create an open AI governance ecosystem, integrating numerous partners for enhanced security and identity management.

AI Agents & GenAI Security: Devvret Rishi on the Risks
Devvret Rishi discusses how AI agents challenge GenAI security, emphasizing the need for dynamic monitoring and governance.

Cursor's Auto-review Balances Agent Autonomy
Cursor's Auto-review feature dynamically manages AI agent autonomy, using a classifier to balance productivity with security risks and minimize user interruptions.

Together AI Locks Down Enterprise Trust
Together AI achieves ISO 27001:2022 certification, enhancing trust and security for its enterprise AI platform and customer data.

Snowflake Bolsters AI Security
Snowflake enhances its platform with new AI security features, including agent identity management and prompt injection protection, to secure enterprise data in the age of autonomous AI.
Databricks curbs AI agent dangers
Databricks Unity Catalog introduces granular control and logging for AI agents, preventing unauthorized actions and providing a complete audit trail.
Securing AI Agents: A New Red Teaming Frontier
A new AI red teaming platform, DTap, and its autonomous agent DTap-Red are introduced to systematically evaluate and secure AI agents across diverse real-world domains.
OpenAI's Safety Playbook for Codex
OpenAI details its robust safety measures for its Codex AI coding agent, emphasizing sandboxing, network controls, and detailed telemetry for secure deployment.

AI Hackers on the Rise: XBOW's Oege de Moor Discusses Autonomous AI
Oege de Moor of XBOW discusses the rise of autonomous AI hackers and their impact on cybersecurity, highlighting rapid AI model advancements.

AI Agents Lack Identity, Risking Enterprise Trust
Enterprises are struggling with the AI agent identity problem, a critical gap in governance and accountability that hinders trust and adoption.

AI Agents: Beyond Chatbots with Open Source
Cedric Clyburn from Red Hat explores the evolution from chatbots to AI agents, detailing the architecture of OpenCopilot and its security implications.
Agentic AI Needs Smarter Guardrails
LangGuard's agentic workflow governance engine, powered by Databricks Lakebase, provides critical runtime control for enterprise AI deployments.

Anthropic's Mythos AI Accessed by Unauthorized Users
Unauthorized users gained access to Anthropic's powerful Mythos AI model, raising security concerns.

Cloudflare Builds the Agentic Cloud
Cloudflare unveils its 'agentic cloud' vision with new tools for building and scaling AI agents, addressing compute, security, and infrastructure needs.
OpenAI Opens AI Cyber Tools to Select Defenders
OpenAI launches 'Trusted Access for Cyber' to equip defenders with advanced AI, prioritizing controlled access and broad ecosystem support.
Databricks Tames Agentic AI
Databricks enhances its AI Gateway to provide unified governance, visibility, and guardrails for complex agentic AI workflows.
ClawGuard Secures LLM Agents
ClawGuard offers a deterministic runtime security framework to prevent indirect prompt injection in LLM agents by enforcing user-confirmed rules at tool-call boundaries.

GitHub's New Game Tests AI Agent Security
GitHub's new Secure Code Game Season 4 challenges developers to hack an AI agent, simulating real-world security risks.

Cloudflare's MCP Security Playbook
Cloudflare outlines its robust security architecture for enterprise-wide adoption of Model Context Protocol (MCP), integrating SASE and developer platforms.

Cloudflare Bolsters Sandbox Security
Cloudflare's new outbound Workers feature provides enhanced security and control for sandboxed AI applications, enabling dynamic authentication and Zero Trust principles.

Secure Agentic AI: Key Takeaways for MCP Servers
Tun Shwe and Jeremy Frenay of Lenses break down the security challenges of MCP servers for agentic AI, offering 5 key rules for secure design.

AI Hacker "Pliny the Liberator" Tests GPT-4 Security
AI security researcher "Pliny the Liberator" demonstrates a novel jailbreaking technique using "tokenades" to manipulate AI models, showcasing the ongoing challenges in AI security.

Anthropic Claude Code leak sparks backlash
Anthropic's Claude Code agent source code was accidentally leaked via an npm package, exposing internal workings and unreleased features.

IBM Expert Details Top 10 Agent Security Risks
IBM Distinguished Engineer Jeff Crume breaks down the OWASP Top 10 vulnerabilities for AI agents, including goal hijacking, tool misuse, and rogue agent behavior.
Databricks Tackles Agentic AI Risks
Databricks enhances its AI Security Framework with 35 new risks and 6 controls for autonomous agent deployment, focusing on memory, planning, and tool usage.

Snowflake Bolsters AI Governance
Snowflake enhances its AI governance capabilities by integrating Bedrock Data into its Horizon Catalog and Cortex AI, addressing critical data classification and control challenges.
Codex Security Ditches SAST Reports
OpenAI's Codex Security agent bypasses SAST reports, focusing on code behavior and intent to find deeper vulnerabilities.

IBM Experts Detail AI Agent Security Imperatives
IBM security leaders Bob Kalka and Tyler Lynch discuss critical security imperatives for AI agents, focusing on accountability, privilege management, and observability.
OpenAI Tackles AI Agent 'Prompt Injection'
OpenAI is adapting its AI security strategy to counter sophisticated prompt injection attacks, treating them as social engineering challenges.

Cloudflare Bolsters AI App Defenses
Cloudflare launches AI Security for Apps, offering threat detection and free endpoint discovery for AI applications, with new custom topic features and expanded partnerships.
OpenAI Buys Promptfoo
OpenAI is acquiring AI security platform Promptfoo to enhance the security, safety, and evaluation features within its Frontier platform for AI coworkers.

OpenAI Details Malicious AI Use in 2026
OpenAI's 2026 malicious AI report reveals how threat actors combine AI with traditional tools and multiple models, informing industry and society on prevention.

Governing Agentic AI by 2026
As agentic AI trends accelerate towards 2026, robust governance frameworks encompassing identity, policy, and enforcement are crucial for safe and ethical autonomous AI deployment.

Veria Labs raises $3.2M
Veria Labs, founded by top US hackers, raises $3.2M seed funding for its AI platform that automates continuous offensive security testing.