Anthropic has introduced a new Code Review feature for its Claude AI, prompting significant backlash from developers due to high token costs and fears that the tool could displace senior engineering roles. The announcement, made in a blog post on Monday, describes the feature as designed to detect complex issues and suggest fixes for bugs. Unlike lighter alternatives, Anthropic stated the new tool optimizes for depth, resulting in significantly higher operational expenses. The feature bills users based on token usage, with estimates ranging from $15 to $25 per pull request, depending on the size and complexity of the code changes. Thariq Shihipar, an Anthropic engineer, explained on social media that the system requires more compute resources, which allows it to identify harder-to-find bugs. However, this capability has not silenced criticism regarding value and affordability. Daniel Ávila, head of AI at the fintech startup Hedgineer, argued that the new feature offers no tangible improvement over existing methods. He noted that while it removes the need for manual triggers used in GitHub Actions, the functional output remains identical. Other developers focused primarily on the pricing structure. Jean P.D. Meijer, a software startup founder, questioned the cost on social media, stating that a single review costs as much as an unlimited subscription to many dedicated code review applications. He expressed confusion over the token pricing strategy. Karan Kendre, a developer at a design startup, challenged the logic behind charging for reviews when the AI is theoretically capable of writing clean code from the start. He suggested the feature might be a revenue-generating tactic layered on top of existing subscription costs. Another user highlighted the potential impact on human wages, noting that pricing a code review at $20 puts a direct monetary tag on a senior developer's role, potentially threatening job security for engineers in that position. This controversy reflects a broader trend of rising expenses in the AI coding tool sector. Billionaire investor Chamath Palihapitiya recently discussed the financial strain on his startup, 8090, which relies heavily on AI tools like Cursor and Anthropic's services. He revealed that AI costs have tripled since November, with the company on track to spend approximately $10 million annually. Palihapitiya noted that while Claude Code was previously a more affordable option compared to Cursor, the overall industry trajectory is leading to unsustainable costs where expenses grow three times faster than revenue. As companies face these inflationary pressures, the industry continues to grapple with balancing advanced AI capabilities against the economic reality for developers.