Claude Desktop silently spawns local VM

// 2h agoNEWS

Claude Desktop silently spawns local VM

Users of the Claude Desktop application report that the app automatically spins up a local virtual machine (VM) consuming up to 1.8 GB of RAM, even for basic chat. Because the application provides no settings or toggles to stop or manage the VM, users must manually kill system services or run scripts to reclaim resources.

// ANALYSIS

Anthropic's decision to silently run a local virtual machine by default prioritizes developer security convenience at the expense of user transparency and resource efficiency.

* Silent overhead: Spawning a local VM by default is excessive for users who only want standard chat functionality and should be strictly opt-in.

* Poor UX design: Forcing users to resort to manual terminal commands or system monitor tools to kill a background VM is a major developer UX oversight.

* Security vs. resource trade-offs: While sandboxing is critical for safe agentic code execution, users should have explicit control over when and how these virtualization resources are allocated.

// TAGS

claude-desktopanthropicvirtualizationhyper-vsandboxuxperformance

DISCOVERED

2h ago

2026-06-10

PUBLISHED

5h ago

2026-06-10

RELEVANCE

8/ 10

AUTHOR

tonyrice

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

TUTORIAL17m ago

Viktor Oddy designs 3D websites on Bolt.new

In a post shared by Bolt.new, creator Viktor Oddy shares a 12-minute step-by-step tutorial on building sophisticated 3D websites within minutes. The workflow demonstrates how to leverage the reasoning and coding capabilities of Anthropic's Claude Opus 4.8 model inside Bolt.new's browser-based development environment to create animated, high-performance web pages without requiring manual design or coding skills.

MODEL34m ago

Anthropic releases Claude Fable 5, a public version of its powerful Mythos AI model equipped with safety guardrails that reroute high-risk queries to Claude Opus 4.8.

Anthropic recently announced the release of Claude Fable 5, a public-facing version of its highly capable, agentic "Mythos-class" AI model. Claude Fable 5 offers significant advances in agentic multi-step task execution, complex reasoning, and coding capabilities, scoring 80.3% on SWE-Bench Pro. To manage risks associated with frontier capabilities, Anthropic has implemented safety classifiers that automatically redirect high-risk queries—such as those related to cybersecurity, biology, or chemistry—to the less powerful Claude Opus 4.8 model. However, the release introduces new enterprise challenges, including a mandatory 30-day data retention policy that overrides existing zero-retention agreements and a heightened need for organizations to implement Zero Trust architectures and update their third-party vendor risk assessments.

MODEL40m ago

Claude Fable 5 impresses Anthropic skeptics

Anthropic has released Claude Fable 5, a "Mythos-class" model designed to autonomously plan, execute, and verify complex, long-running agentic tasks. To manage safety risks, Fable 5 automatically routes sensitive prompts to Claude Opus 4.8 and requires a 30-day safety-monitoring data retention period.