Did OpenAI Leverage Competitor Claude’s Coding Tools for GPT-5 Training? Anthropic Suggests It’s Possible

The AI industry is witnessing a fierce competition as OpenAI prepares to launch its next-generation model, GPT-5, while Anthropic accuses it of misusing its Claude models. Anthropic claims OpenAI’s engineers utilized Claude’s coding tools for internal testing, violating terms of service. A spokesperson pointed out that instead of accessing Claude through standard user interfaces, OpenAI allegedly connected directly to Anthropic’s developer APIs, allowing extensive testing in various areas, including coding and sensitive topics. Despite cutting off OpenAI’s access, Anthropic will still permit API access for benchmarking and safety evaluations. OpenAI expressed disappointment but acknowledged industry standards in evaluating AI systems. This incident highlights the intense rivalry in the AI space as companies strive for innovation and ethical practices. As the release of GPT-5 approaches, the implications of these developments emphasize the significance of adhering to commercial agreements in the competitive AI landscape.

Source link