San Francisco-based AI lab Anthropic is reportedly exploring designing its own custom chips, a strategic move that could reshape the global AI supply chain. According to sources familiar with the matter, the company behind the Claude AI model is considering the move to navigate the persistent global shortage of advanced processors required to train and run powerful AI systems.
Quick Facts
- Exploring in-house AI chip design.
- Plans are in early, uncommitted stages.
- Mirrors similar efforts by Meta and OpenAI.
Navigating the AI Compute Bottleneck
Anthropic’s consideration comes as the demand for its AI models skyrockets. The company stated earlier this week that its run-rate revenue has surpassed $30 billion, a significant jump from about $9 billion at the end of 2025, signaling intense pressure on its computing infrastructure.
The plans are still preliminary, with sources noting that no dedicated team has been formed and the company may ultimately continue to rely solely on third-party suppliers.
A Costly Bet on Vertical Integration
Currently, Anthropic leverages a mix of hardware to power Claude, including Google’s Tensor Processing Units (TPUs) and chips from Amazon Web Services. The company recently reinforced its partnership with Google through a long-term deal that also involves Broadcom, a key designer of the TPUs.
A move into custom silicon is a major undertaking, with industry estimates placing the cost of designing an advanced AI chip at roughly half a billion dollars. This potential pivot reflects a growing trend among tech giants like Meta and OpenAI, who are also pursuing in-house chip development to gain a competitive edge and secure their compute needs.
Implications for the MENA AI Ecosystem
This potential shift by a foundational AI model provider carries significant weight for the MENA region. As countries like the UAE and Saudi Arabia pour billions into becoming global AI hubs and acquiring vast quantities of GPUs, the global supply chain for AI hardware is a critical dependency.
If major AI labs begin to internalize their chip design, it could further tighten the market for state-of-the-art processors available to other players. For MENA startups and enterprises building on models like Claude, this signals a future where access to the most powerful AI may become even more competitive. It underscores the strategic importance of the region’s investments in data centers and computing infrastructure, and may accelerate efforts to forge direct partnerships with hardware manufacturers.
About Anthropic
Anthropic is an AI safety and research company dedicated to building reliable, interpretable, and steerable AI systems. Founded by former members of OpenAI, the company is known for its family of large language models named Claude, which are designed to be helpful, harmless, and honest.
Source: Tech in Asia
