All news with #amazon bedrock tag

🚀 Anthropic's Claude Sonnet 4.5 is now available in Amazon Bedrock within AWS GovCloud (US‑West and US‑East) via US‑GOV Cross‑Region Inference. The model emphasizes advanced instruction following, superior code generation and refactoring judgment, and is optimized for long‑horizon agents and high‑volume workloads. Bedrock adds an automatic context editor and a new external memory tool so Claude can clear stale tool-call context and store information outside the context window, improving accuracy and performance for security, financial services, and enterprise automation use cases.

🧰 Amazon Bedrock AgentCore Runtime now supports two deployment methods: direct code-zip upload and container-based deployment. Developers can use drag-and-drop code-zip uploads for rapid prototyping or opt for container images when they need custom runtime configurations and dependencies. The serverless, model-agnostic runtime is designed to scale for production while maintaining enterprise security. This capability is available across nine AWS Regions with consumption-based pricing and no upfront costs.

🔒 The Model Context Protocol (MCP) Proxy for AWS is now generally available, offering a client-side proxy that lets MCP clients connect to remote, AWS-hosted MCP servers using AWS SigV4 authentication. It supports agentic development tools such as Amazon Q Developer CLI, Kiro, Cursor, and agent frameworks like Strands Agents, and interoperates with MCP servers built on Amazon Bedrock AgentCore Gateway or Runtime. The open-source Proxy includes safety controls (read-only mode), configurable retry logic, and logging for troubleshooting, and can be installed from source, via Python package managers, or as a container to integrate with existing MCP-supported tools.

🚀 Amazon expanded availability of TwelveLabs Pegasus 1.2 to US East (Ohio), US West (N. California), and Europe (Frankfurt) via Amazon Bedrock. Pegasus 1.2 is a video-first language model optimized for long-form video understanding, video-to-text generation, and temporal reasoning across visual, audio, and textual signals. The regional rollout brings the model closer to customers' data and end users, reducing latency and simplifying deployment architectures. Developers can now build enterprise-grade video intelligence applications in these regions.

🔐 Amazon Bedrock AgentCore Browser now previews Web Bot Auth, a draft IETF protocol that cryptographically identifies AI agents to websites. The feature automatically generates credentials, signs HTTP requests with private keys, and registers verified agent identities to reduce CAPTCHA interruptions and human intervention in automated workflows. It streamlines verification across major providers such as Akamai, Cloudflare, and HUMAN Security, and is available in nine AWS Regions on a consumption-based pricing model with no upfront costs.

🎥 TwelveLabs' Marengo Embed 3.0 is now available on Amazon Bedrock, providing a unified video-native multimodal embedding that represents video, images, audio, and text in a single vector space. The release doubles processing capacity—up to 4 hours and 6 GB per file—expands language support to 36 languages, and improves sports analysis and multimodal search precision. It supports synchronous low-latency text and image inference and asynchronous processing for video, audio, and large files.

🖼 Amazon Bedrock now offers four new image-editing tools in Stability AI Image Services: Outpaint, Fast Upscale, Conservative Upscale, and Creative Upscale. These additions expand the platform's Edit, Upscale, and Control capabilities, enabling creators to perform targeted edits and resolution enhancements with greater precision. The tools are accessible via the Bedrock API and are initially supported in US West (Oregon), US East (N. Virginia), and US East (Ohio).

🌐 Web Grounding is now generally available as a built-in tool for Nova models, usable today with Nova Premier via the Amazon Bedrock tool use API. It retrieves and incorporates publicly available information with citations to support responses, enabling a turnkey RAG solution that reduces hallucinations and improves accuracy. Cross-region inference makes the tool available in US East (N. Virginia), US East (Ohio), and US West (Oregon). Support for additional Nova models will follow.

🚀 Amazon announces general availability of Amazon Nova Multimodal Embeddings, a unified embedding model designed for agentic RAG and semantic search across text, documents, images, video, and audio. The model handles inputs up to 8K tokens and video/audio segments up to 30 seconds, with segmentation for larger files and selectable embedding dimensions. Both synchronous and asynchronous APIs are supported to balance latency and throughput, and Nova is available in Amazon Bedrock in US East (N. Virginia).

🎞️ Amazon Bedrock Data Automation now supports AVI, MKV, and WEBM video formats and the AV1 and MPEG-4 Visual (Part 2) codecs, expanding coverage for archival, multi-track, and web-based videos. The service also delivers up to 50% faster image processing to accelerate extraction of visual insights. BDA is available in eight AWS Regions, enabling organizations to process native formats and streamline GenAI workflows.

🔒 Amazon announced that Amazon Nova models now support customizable content moderation settings for approved business use cases that require processing or generating sensitive content. Organizations can adjust controls across four domains—safety, sensitive content, fairness, and security—while Amazon enforces essential, non-configurable safeguards to protect children and preserve privacy. Customization is available for Amazon Nova Lite and Amazon Nova Pro in the US East (N. Virginia) region; customers should contact their AWS Account Manager to confirm eligibility.

🔒 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.

🔐 Amazon Bedrock now automatically enables serverless foundation models in each AWS Region, removing the prior per-model enablement step and retiring the Model Access page and PutFoundationModelEntitlement IAM permission. Access is managed through standard AWS controls—IAM and Service Control Policies (SCPs)—so account- and organization-level governance remains intact. Existing model restrictions enforced by IAM or SCPs continue to apply, and previously enabled models are unaffected. Administrators should transition to scoped IAM/SCP policies and patterns such as wildcards and NotResource denies to maintain least-privilege control.

🔓 Amazon Bedrock now automatically enables access to all serverless foundation models by default in all commercial AWS regions. This removes the prior manual activation step and lets users immediately use models via the Amazon Bedrock console, AWS SDK, and features such as Agents, Flows, and Prompt Management. Anthropic models remain enabled but require a one-time usage form before first use; completing the form via the console or API and submitting it from an AWS organization management account will enable Anthropic across member accounts. Administrators continue to control access through IAM policies and Service Control Policies (SCPs).

🚀 Claude Haiku 4.5 is now available in Amazon Bedrock, offering near-frontier performance comparable to Claude Sonnet 4 while reducing cost and improving inference speed. The model targets latency-sensitive and budget-conscious deployments, excelling at coding, computer use, agent tasks, and vision-enabled workflows. Haiku 4.5 supports global cross-region inference and is positioned for scaled production use; consult Bedrock documentation, the console, and pricing pages for region and billing details.

🚀 Amazon Bedrock has expanded regional access to several foundation models, adding DeepSeek-V3.1, OpenAI open-weight models (20B, 120B), and multiple Qwen3 variants. The update makes DeepSeek-V3.1 and Qwen3 Coder-480B available in US East (Ohio) and Asia Pacific (Jakarta), and brings OpenAI open-weight and additional Qwen models to US East (Ohio), Europe (Frankfurt), and Asia Pacific (Jakarta). Customers can deploy these models locally to meet data residency needs, reduce latency, and enable faster AI-powered experiences.

🔐 Amazon Bedrock AgentCore Identity centralizes and secures identities and credentials for AI agents, integrating with existing identity providers such as Amazon Cognito to avoid user migration and rework of authentication flows. It provides a token vault encrypted with AWS KMS, native AWS Secrets Manager support, and orchestrates OAuth 2.0 flows (2LO and 3LO). Declarative SDK annotations and built-in error handling simplify credential injection and refresh workflows, helping teams deploy agentic workloads securely at scale.

🔐 Amazon Bedrock AgentCore is now generally available, offering an agentic platform to build, deploy, and operate agents securely at scale without infrastructure management. It adds VPC support, extended eight-hour runtimes, session isolation, and initial Agent-to-Agent (A2A) protocol support. Identity, Gateway, and Observability features provide secure authorization, tool discovery, and CloudWatch/OTEL monitoring.

🚀 Amazon ElastiCache now offers vector search generally available with Valkey 8.2, enabling indexing, searching, and updating billions of high-dimensional embeddings from providers such as Amazon Bedrock, Amazon SageMaker, Anthropic, and OpenAI with microsecond latency and up to 99% recall. Key use cases include semantic caching for LLMs, multi-turn conversational agents, and RAG-enabled agentic systems to reduce latency and cost. Vector search runs on node-based clusters in all AWS Regions at no additional cost, and existing Valkey or Redis OSS clusters can be upgraded to Valkey 8.2 with no downtime.

⚠️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.