Last chance: 5 steps to secure your AI supply chain

By 2029, 60% of enterprise AI breaches will result from compromised third-party dataset or model providers, rather than software vulnerabilities, according to Gartner®.

This report walks through 5 steps to secure your AI supply chain, but it won’t be available much longer! Download it before it’s gone.

Local enterprise AI, no compromises: Anaconda Desktop + NVIDIA DGX Spark

Anaconda CEO David DeSanto built a working local AI agent in under 3 minutes running entirely on hardware sitting on his desk, no cloud required.

Part 1 of his series covers the full setup: from unboxing NVIDIA DGX Spark to a running Model API Server and first agent in Anaconda Agent Studio. A useful primer on what’s possible when local and enterprise AI stop being trade-offs.

Get free access to AI Agents in Action, Second Edition

The second edition of AI Agents in Action is out and free access is available now.

Updated for 2026, it covers the frameworks, orchestration patterns, and real-world workflows teams are using to build and ship production-grade AI agents today. Free access won’t last.

Your Python workflow, now AI-ready

Two new experimental features in Anaconda Desktop Beta close the gap between your Python environments and the AI tools you already use.

Anaconda Agent Studio lets you build, test, and run agents without leaving your existing Python workflow. Anaconda MCP gives tools like Claude Code and Cursor direct access to your conda environments.

NVIDIA Nemotron 3 Ultra is now available from Anaconda

NVIDIA’s largest open frontier reasoning model, with 550 billion parameters, 1M token context window, built for agentic workflows, is now available on Anaconda Platform.

Pair it with Nemotron 3 Nano for a complete agentic stack: Ultra handles orchestration, long-horizon planning, and complex reasoning. Nano handles high-volume tool calls and execution. Both run inside a single project structure with unified observability

OpenAI is weighing drastic price cuts as competition with Anthropic intensifies

According to the Wall Street Journal, OpenAI is considering significant token price reductions as Anthropic gains ground among developers, a notable move for a company already losing billions on compute.

As foundation model pricing commoditizes, the value shifts to what sits around the model: data quality, reproducible environments, and governed infrastructure. A conversation worth following.