< ciso
brief />
Tag Banner

All news with #bedrock guardrails tag

27 articles · page 2 of 2

DNS Exfiltration and RCE Risk in AI Code Sandboxes

🔒 Researchers disclosed that Amazon Bedrock AgentCore Code Interpreter's sandbox mode permits outbound DNS queries, enabling attackers to create bidirectional command-and-control channels and exfiltrate data via DNS despite a "no network access" setting. BeyondTrust rated the issue 7.5/10 and recommends migrating critical workloads to VPC mode and using a Route53 DNS Firewall. Administrators should audit IAM roles and inventory active interpreters immediately.
read more →

AgentCore Policy Controls in Amazon Bedrock Now Available

🔒 Amazon has made Policy in Bedrock AgentCore generally available, providing centralized, fine-grained controls for agent-to-tool interactions. Teams can author policies in natural language that AWS converts into Cedar and stores in a policy engine attached to an AgentCore Gateway, which intercepts traffic and evaluates requests before allowing or denying access. Operating outside agent code, this lets security, compliance, and operations enforce access rules and validate inputs without modifying agents, improving governance and visibility across deployments.
read more →

Bedrock Guardrails: Natural-Language Test Generation

🧪 Amazon Web Services has added natural-language test Q&A generation to Automated Reasoning checks in Amazon Bedrock Guardrails. The capability generates up to N test Q&As from input documents to accelerate creating and validating formal verification policies. Automated Reasoning checks apply formal methods to detect correct model outputs and report up to 99% accuracy in identifying correct responses and reducing hallucinations. The feature is available in multiple US and EU Regions and accessible via the Bedrock console and Python SDK.
read more →

Amazon Bedrock Guardrails Expand Code-Related Protections

🔒 Amazon Web Services expanded Amazon Bedrock Guardrails to cover code-related use cases, enabling detection and prevention of harmful content embedded in code. The update applies content filters, denied topics, and sensitive information filters to code elements such as comments, variable and function names, and string literals. The enhancements also include prompt leakage detection in the standard tier and are available in all supported AWS Regions via the console and APIs.
read more →

Encoding-Based Attack Protection with Bedrock Guardrails

🔒 Amazon Bedrock Guardrails offers configurable, cross-model safeguards to protect generative AI applications from encoding-based attacks that attempt to hide harmful content using encodings such as Base64, hexadecimal, ROT13, and Morse code. It implements a layered defense—output-focused filtering, prompt-attack detection, and customizable denied topics—so legitimate encoded inputs are allowed while attempts to request or generate encoded harmful outputs are blocked. The design emphasizes usability and performance by avoiding exhaustive input decoding and relying on post-generation evaluation.
read more →

Indirect Prompt Injection Poisons Agents' Long-Term Memory

⚠️This Unit 42 proof-of-concept shows how an attacker can use indirect prompt injection to silently poison an AI agent’s long-term memory, demonstrated against a travel assistant built on Amazon Bedrock. The attack manipulates the agent’s session summarization process so malicious instructions become stored memory and persist across sessions. When the compromised memory is later injected into orchestration prompts, the agent can be coerced into unauthorized actions such as stealthy exfiltration. Unit 42 outlines layered mitigations including pre-processing prompts, Bedrock Guardrails, content filtering, URL allowlisting, and logging to reduce risk.
read more →

Adapting Enterprise Risk Management for Generative AI

🛡️ This post explains how to adapt enterprise risk management frameworks to safely scale cloud-based generative AI, combining governance foundations with practical controls. It emphasizes the cloud as the foundational infrastructure and identifies differences from on‑premises models that change risk profiles and vendor relationships. The guidance maps traditional ERMF elements to AI-specific controls across fairness, explainability, privacy/security, safety, controllability, veracity/robustness, governance, and transparency, and references tools such as Amazon Bedrock Guardrails, SageMaker Clarify, and the ISO/IEC 42001 standard to operationalize those controls.
read more →