aiagentnews.io

Monday 16 November

Amazon Bedrock lets teams set per-tenant answer instructions on managed knowledge bases

AI-generated summary Amazon Bedrock now lets teams set per-tenant answer instructions on managed knowledge bases. Developers can use four fields including persona and answerPolicy to manage tone and format per tenant.

Earlier:

Amazon Bedrock Managed Knowledge Base adds tabular search for CSV and Excel files

AI-generated summary Amazon Bedrock Managed Knowledge Base now includes tabular search for CSV and Excel files to help agents answer questions requiring calculations across rows. The feature works through the existing AgenticRetrieveStream request and supports files from connectors such as Amazon S3, Microsoft SharePoint and Atlassian Confluence.

Also covered byCloudNinjasZeniteq
Earlier:

Agents are sessions, not requests: size GKE node pools for AI agent sessions- Google Developers Blog

AI-generated summary Google Developers Blog discusses sizing GKE node pools for AI agent sessions based on GitHub Copilot agent traces and usage patterns. Agent Substrate 1.0 is generally available on GKE following the CNCF donation of gVisor, Agent Substrate, and OpenRL in November 2026.

Also covered byaiagentnews.io
Earlier:

Claude now builds websites inside canvas graphics, and the HTML stays clickable

AI-generated summary Anthropic's AI assistant Claude can now use a browser technology called HTML-in-Canvas to build web pages inside canvas graphics. Felix Rieseberg brought the feature to wider attention, and users need Chrome or Edge or Claude's desktop apps to view it.

Also covered byThe Decoderaiagentnews.ioThe Register
Earlier:

Noeri Secures $110M in Seed Funding for Customized Enterprise AI Software

Noeri Secures $110M in Seed Funding for Customized Enterprise AI Software

AI-generated summary The newly formed artificial intelligence startup Noeri secured $110M in a seed funding round led by Menlo Ventures to build customized enterprise software. The financing round included participation from Madrona Ventures, NEA, Cisco, and Snowflake for the tools designed by founders Yejin Choinka and Carlos Guestrin.

Also covered byaiunderstanding.orgSiliconANGLEaiagentnews.io

Anthropic says the key to agent memory is knowing what to forget, not how much to store

AI-generated summary Anthropic studied how Claude retains knowledge across sessions and concluded that selective notes outperform exhaustive archives. Anthropic says that a useful agent memory depends on keeping the right information and reviewing memories to correct errors.

Also covered byaiagentnews.io
Earlier:

Claude Code introduces experimental Code Mode for tool invocation

Claude Code introduces experimental Code Mode for tool invocation

AI-generated summary Anthropic added Code Mode in the v2.1.323 release of Claude Code on November 12. The model writes a single JavaScript script for tool invocation instead of sequential calls.

Also covered byDigestXaiagentnews.ioDigIntoAI
Earlier:

Dreamteam Raises $53 Million for an AI-Native CRM With Specialized Sales Coworkers

AI-generated summary Dreamteam has raised $53 million from Accel, Together Fund, and Premji Invest to scale an AI-native customer relationship management platform. The platform utilizes a shared memory system and four specialized assistants to capture context from customer interactions and sales workflows.

Also covered byFusion42TechCrunchaiagentnews.io

Qwen4-Max: New model architecture, phased delivery for long tasks

On a white background, layered navy and gray-blue 3D blocks on the left connect via teal lines to four gray-blue rounded rectangles above, below, left, and right. The label below reads ‘New Architecture’. On the right, five rounded rectangles in progressively darker teal sit horizontally among pale teal bands, with three small dots and a teal circle containing a white check mark. The label below reads ‘Live Delivery’.

AI-generated summary Alibaba unveiled Qwen4-Max, a model featuring a sparse mixture-of-experts architecture and a context limit of 1 million tokens. The company reports that Qwen4-Max uses phased delivery for long tasks to let developers check results early and change direction as work progresses.

Also covered byCode404ZeroTekediaaiagentnews.io

Cyber startup Corma goes from stealth to unicorn in just three months with $125 million Series A

AI-generated summary Cybersecurity startup Corma has raised $125 million in a Series A funding round at a valuation of $1 billion. Corma develops an autonomous workforce for cybersecurity powered by a proprietary foundation model designed for defensive security.

Also covered byaiunderstanding.orgHuggingNewsaiagentnews.io

OpenAI Agents SDK v0.24.0-alpha.8 adds managed tool approvals and Realtime voice

AI-generated summary The release of the OpenAI Agents SDK introduces managed tool approvals and Realtime voice capabilities for developers. The alpha update also adds checkpoint support for pending runs and includes breaking changes for existing code or saved state.

Earlier:

Samsung's 'Galaxy Claw' is no longer a mystery, and it looks pretty ambitious

AI-generated summary Samsung is reportedly working on a new Galaxy AI experience called Galaxy Claw that could perform everyday tasks by connecting to apps and services on the user device. References to the feature were found in work-in-progress code within the Personal Agent Core app, though these capabilities may not ship.

Also covered byDigital TrendsZDNETaiagentnews.io

Sunday 15 November

Uber's Secure AI Framework Governs Autonomous AI Agents Across Their Lifecycle

AI-generated summary InfoQ reports that Uber developed a framework to govern autonomous AI agents across their lifecycle to address limitations in traditional identity and API controls. Tristan G. and Pratik Dhanave shared observations on the framework regarding agent constraints, costs, and implementation challenges.

Also covered byaiagentnews.ionews.lavx.huamartuvshin.com