Latest Insights

Overview of AI Security

Foundational AI risks (including AI agents) and the evolution of security strategies.

• Overview of AI Security

Securing Agents

Protecting autonomous workflows and agent interactions.

• Securing Agents

Securing Products

Integrating security into product lifecycles.

• Securing Products

Securing Models

Ensuring integrity and robustness across the model lifecycle.

• Securing Models

Securing Infrastructure

Hardening the foundation through silicon-to-software isolation.

• Securing Infrastructure

Securing Data

Protecting data and privacy throughout the AI process.

• Securing Data

Testing and Assurance

Validating safety through empirical discovery and defense.

• Testing and Assurance

Detection and Response

Detecting threats to AI, threat actors using AI, and responding to AI incidents.

• Detection and Response

Google Cloud Security Blog
October 2026
How Google Does It: Using agents to help find, track, and fix vulnerabilities

How Google uses autonomous AI agents and automated patching pipelines to find, triage, and remediate software vulnerabilities at machine speed.

• Securing Infrastructure
• Testing and Assurance
• Detection and Response

Google Public Policy
September 2026
Securing the Agentic Era: Our Commitments and Proposed Action Blueprint

Google’s policy commitments and action blueprint for securing AI agents, modernizing digital infrastructure, and advancing ecosystem standards.

• Securing Agents
• Overview of AI Security
• Securing Infrastructure

Google Security Blog
August 2026
The Evolving Role of the Red Team in the Era of Agentic Security

How AI red teaming is evolving to evaluate and harden autonomous agents and foundation models against emerging threats.

• Testing and Assurance
• Securing Agents
• Securing Models

Google Research
October 2025
Google’s Approach to Protecting Privacy in the Age of AI

A framework for reasoning about privacy in AI and approaches to reducing privacy risks.

• Overview of AI Security
• Securing Data

Google Research
September 2025
Our Approach to Protecting AI Training Data

Application of best practices in data protection to AI systems.

• Securing Data
• Securing Models

Google Research
September 2025
Security Assurance in the Age of Generative AI

Practical approaches to implementing Google’s Secure AI Framework.

• Testing and Assurance
• Securing Models

Google Research
May 2025
Google’s Approach for Secure AI Agents

An aspirational framework advocating for a hybrid, defense-in-depth strategy.

• Securing Agents
• Securing Products
• Securing Infrastructure

Google Cloud Security Community
February 2025
Securing AI Supply Chain: Like Software, Only Not

Why securing the AI and model supply chain differs from traditional software supply chains, and how to mitigate model lifecycle risks.

• Securing Models
• Securing Products

SAIF.google
October 2024
Secure AI Development Primer

An introduction to the AI development lifecycle, and the risks that can emerge at each stage.

• Overview of AI Security
• Securing Models
• Securing Products

External webinar
June 2024
Protecting your AI: Strategies for Securing AI Systems

Google security leaders share best practices for safeguarding AI environments.

• Securing Products
• Securing Models
• Overview of AI Security

Google DeepMind
May 2024
Introducing the Frontier Safety Framework

Proactive protocols for identifying severe capabilities and implementing security and safety mitigations across frontier AI models.

• Securing Models
• Testing and Assurance

Google Research
May 2024
Securing the AI Software Supply Chain

Adapting established approaches to AI supply chain security without reinventing the wheel.

• Securing Products
• Securing Models
• Securing Infrastructure

Google Cloud
March 2024
Best Practices for Securely Deploying AI on Google Cloud

Practical guidance for hardening cloud infrastructure, securing model deployment pipelines, and protecting AI workloads.

• Securing Infrastructure
• Securing Models
• Securing Data

Google Cloud
February 2024
How AI Can Reverse the Defender’s Dilemma

Tilting the scales to give defenders a decisive advantage over attackers.

• Detection and Response
• Overview of AI Security

Google Research
January 2024
AI-Powered Patching: The Future of Automated Vulnerability Fixes

Lessons from our experience leveraging AI to scale our ability to fix bugs.

• Detection and Response
• Testing and Assurance