Anthropic denies intentional slowdown of Claude Code

2026-04-24 · Source: Dataconomy · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, quick

Summary

Anthropic acknowledged and addressed user complaints regarding declining performance of its AI tool, Claude Code, over several weeks. The company explicitly denied any intentional "nerfing" or degradation of the underlying model, attributing the issues to product-level adjustments. Following a detailed review, Anthropic identified three specific problems: a change to Claude Code’s default thinking level, a bug from a cache-optimization tweak, and an adjustment to reduce verbosity. As of April 20, the company announced these issues were fixed, and it has implemented measures like increased internal usage of public builds, enhanced code review tools, and stricter controls on system prompt changes to prevent recurrence. This resolution comes amidst user speculation about intentional slowdowns, which Anthropic firmly rejected.

Key takeaway

For engineering teams managing AI product deployments, promptly investigating and transparently communicating about performance degradation is crucial. Your users are a vital feedback loop; ignoring their reports can erode trust and lead to negative speculation. Implement robust internal testing and strict change management for product-level adjustments to prevent unintended performance impacts, especially for critical tools like code assistants.

Key insights

Product-level adjustments, not core model degradation, caused Claude Code's performance issues.

Principles

• Transparency builds trust during performance issues.
• User feedback is critical for product quality assurance.

Method

Anthropic identified issues through user reports and internal review, then implemented fixes including resetting default parameters, patching bugs, and enhancing internal testing and control mechanisms.

In practice

• Monitor user feedback channels closely.
• Implement strict change controls for system prompts.
• Increase internal testing of public builds.

Topics

• Anthropic
• Claude Code
• AI Model Performance
• User Feedback
• Product Development Issues

Best for: Machine Learning Engineer, CTO, VP of Engineering/Data, Software Engineer, AI Engineer, AI Product Manager

Related on AIssential

• Anthropic tested removing Claude Code from the Pro plan — Evolving user engagement with AI agents necessitates dynamic adjustments to service plans and resource allocation.
• Anthropic Traces Six Weeks of Claude Code Quality Complaints to Three Overlapping Product Changes — Overlapping product-layer changes, not model weights, caused Claude Code's six-week quality decline.
• Mystery solved: Anthropic reveals changes to Claude's harnesses and operating instructions likely caused degradation — Product-layer changes, not model degradation, caused Claude's perceived performance issues.
• Anthropic confirms Claude Code problems and promises stricter quality controls — Perceived AI quality drops often stem from infrastructure or tooling changes, not core model regressions.
• How Anthropic’s product team moves faster than anyone else | Cat Wu (Head of Product, Claude Code) — Rapid iteration and clear product taste are paramount for PMs in the fast-evolving AI product development landscape.
• Anthropic’s Code with Claude showed off coding’s future—whether you like it or not — LLMs are rapidly automating software development, shifting human roles from writing to oversight, despite emerging concerns.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Dataconomy.