Amazon’s Claude Code Limitations Frustrate Employees, Affecting Cloud Sales

The relationship between Amazon and Anthropic, a leading AI startup, is dynamic but marked by internal friction. Despite being one of Anthropic’s largest investors, Amazon restricts employees from using Anthropic’s AI model, Claude Code, in production without approval. This policy has stirred discontent among workers.

A Shift in Internal Policy

Internal guidance unveiled in late 2022 directed Amazon teams to favor Kiro, the company’s proprietary AI coding assistant, over Claude Code. This move prompted widespread discussions among employees. Approximately 1,500 Amazon workers have voiced their support for utilizing Claude Code, citing the limitations of Kiro.

Concerns Around Cloud Sales

Many engineers responsible for promoting AWS Bedrock, Amazon’s platform offering access to third-party AI services, find themselves in a precarious position. They grapple with the challenge of selling a product they cannot formally use during their development work.

• Employees expressed concerns that skepticism could arise among customers, asking why they should trust an unapproved tool.
• A sales engineer commented on the irony of promoting Claude Code without internal support.
• Employees emphasized the need to use Claude Code for effective sales demonstrations.

Performance and Productivity Issues

Debates among engineers have also highlighted potential productivity hurdles linked to promoting Kiro. Some argue that Claude Code outperforms Kiro, raising concerns about forcing employees to adopt a less effective tool.

• Critics question Kiro’s innovation potential compared to its rivals.
• One employee noted that without genuine competitiveness, Kiro’s success may rely solely on enforced usage.

The Internal Response

Despite the pushback, Amazon maintains a strategic partnership with Anthropic. Amazon’s spokesperson indicated that while Claude Code is not explicitly banned, the company has implemented strict requirements for tools used in production. A process exists for seeking exceptions, but employees have voiced concerns about transparency.

Frustration has grown among staff, particularly as earlier indications suggested that Claude Code might meet the necessary security and legal standards. However, this language was later revised, leading to additional discontent.

Employee Sentiments

Sentiments among employees reflect a clear demand for formal support of Claude Code. With increased internal endorsements, many believe that recognizing Claude Code could enhance productivity and morale.

As Amazon continues to refine its AI strategy, the ongoing debate over Claude Code underscores the complexities of balancing internal policies with external partnerships.