Legora raises $550 million to expand its legal AI platform
The new financing backs a legal AI platform built around request handling, reusable instructions and lawyer review rather than a new foundation model.
AI Agents
Agent Skills moves to AAIF with a roadmap for shared governance
Anthropic is transferring Agent Skills to AAIF. The format remains open, while client permissions, dependencies and workflow testing still matter.
Blacksmith backs CI capacity as coding agents increase testing demand
Blacksmith’s Series C funds CI infrastructure. Its documentation separates GitHub runner access from a private beta for recording evidence about code changes.
Firstwork raises $15 million for frontline workforce AI
Firstwork connects document checks, portal tasks and credential renewals. Its investor describes customer approved rules, while product pages include automated restrictions and clearance.
Opply raises $7.5 million for AI procurement
Opply's funding backs procurement agents for consumer brands. Its current services retain customer approvals, while inventory and production planning remain on the roadmap.
Daytona raises $60 million for its agent execution cloud
Daytona's new funding backs agent infrastructure. Its current docs separate container storage, VM memory, network rules and credential scope.
Pigment raises $250 million for AI business planning
Pigment's Series E backs its AI planning platform. Current tools build models and analyze data, while the Planner Agent remains on the roadmap.
OpenAI acquires Orchestra Bio for drug development workflows
Orchestra brings drug program coordination to OpenAI’s life sciences plans.
Hyper brings cloud browsing to its stable app
Hyper 1.11.0 enables cloud browsing for stable users, with desktop takeover, temporary sessions and limits on mobile sign in.
Cursor adds cloud agent canvases and an environment API
Cursor’s October 16 update adds shared canvases, a beta Environments API and support for 300 repositories, with sharing and access conditions for teams.
OpenCode 2.2.6 adds worktree runs and clearer review errors
OpenCode 2.2.6 clarifies review loading failures, adds scripted worktree runs and waits for current-prompt background jobs. Here are the limits to check.
Meta releases Muse Spark Open with gated weights
Meta’s October 15 release offers BF16 and NVFP4 weights. Mac results vary by checkpoint and setup, while the license adds specific service conditions.
Standard Bots expands AI tools for factory robot operators
Three skills target setup and quality checks.
Figma opens Code layers and Specs for teams building with AI
Figma’s October update brings running prototypes and shared specifications onto the canvas, with rollout, credit and code visibility limits to check.
Harmonic raises $50 million to expand controls on workplace AI
Harmonic’s Series B funds workplace AI controls. Its product tiers separate monitoring from agent permissions, with prompt injection blocking still planned.
Suki announces $90 million financing for healthcare AI agents
Suki plans to expand healthcare AI, integrations and its partner business.
Amex publishes intent framework for AI shopping agents
The framework would give issuers authenticated purchase instructions before an AI agent completes a transaction.
Arena brings Max routing to Agent Mode
Model routing expands to longer tasks. The test dates matter.
Melius announces $25 million in total funding for creative agents
Melius reports cumulative funding for its creative tools. Editable workflows, generation credits and planned features shape the practical evaluation.
Space Bunny’s MiniMax identity and changing access routes
MiniMax identified Space Bunny as an early M3.1 Flash version on October 19. The earlier free-access listings should not be treated as current guarantees.
Sesame gives voice agents their own computers
Sesame adds computer access to its voice assistant preview. Here is what users can try and how Google permissions and conversation controls differ.
OpenCode 1.19 makes its updater manual
OpenCode changes its built in V1 updater and adds a deliberate path to V2. Source inspection reveals what users should check before switching.
LocalLightChat 1.0 adds agents and retrieval to its local AI interface
LocalLightChat adds agent tools, MCP, retrieval and image editing, with practical requirements for connected models, deployment and migration.
GitHub gives Copilot admins finer control over activity logs
GitHub announces granular Copilot telemetry controls while earlier VS Code releases reveal why deployment checks matter.