Anthropic Confirms Claude Code Quality Problems and Promises Stricter Controls

Anthropic acknowledged quality problems with Claude Code after developers complained about declining performance and accused the company of secretly downgrading its Opus 4.6 model, while the newer Opus 4.7 drew mixed reviews over bugs and higher token costs. The AI company identified three sources of error and promised stricter quality controls as it grapples with surging demand that has strained compute capacity, caused a security breach exposing internal files, and led to confusion when Claude Code was removed from some $20 Pro plan subscriptions as part of a limited test. The troubles come as Anthropic faces a potential $800 billion IPO and intensifying competition from OpenAI, whose executives have publicly criticized Anthropic as elitist and accused it of overstating revenue while courting frustrated developers. Despite the operational challenges spanning product quality, pricing, security and capacity over two months, Anthropic's revenue has tripled to $30 billion this year and demand remains strong as the company pushes enterprise clients toward usage-based pricing.

Anthropic's Cascade of Product and Operational Problems

• ▪Anthropic's newest model Opus 4.7 delivered major benchmark gains but drew mixed public reception
• ▪Some users discovered on Tuesday that Claude Code was no longer available on Anthropic's $20 per month Pro plan
• ▪Some developers accused Anthropic of quietly downgrading its flagship model Opus 4.6 due to perceived performance declines
• ▪Some users of Anthropic's Opus 4.7 complained of higher token costs, bugs and uneven performance
• ▪Anthropic said the removal of Claude Code from the Pro plan was part of a limited test affecting a small share of users
• ▪A software update at Anthropic accidentally exposed internal Claude Code files to outsiders

Competitive Dynamics Between Anthropic and OpenAI Ahead of IPOs

• ▪OpenAI CEO Sam Altman accused Anthropic of fear-based marketing in a podcast appearance this week
• ▪OpenAI chief revenue officer Denise Dresser alleged in a leaked memo that Anthropic had overstated its revenue run rate by billions
• ▪OpenAI engineers openly mocked Anthropic on social media as Anthropic grappled with user backlash from its Claude Code pricing confusion
• ▪Anthropic is facing a potential IPO that could value the company near $800 billion
• ▪A leaked memo from OpenAI chief revenue officer Denise Dresser blasted Anthropic as elitist
• ▪OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei used to signal that the AI race should have multiple winners

Business Performance Despite Technical Challenges

• ▪Anthropic's revenue has tripled to $30 billion this year
• ▪Anthropic's standoff with the Pentagon helped drive a spike in usage that briefly sent Claude to the top of the U.S. App Store

Perspective of Frustrated Claude developers

• ▪Developers believe Anthropic quietly downgraded Claude Opus 4.6 without public disclosure or explanation