TheChek News Flow

The Takeaway

Anthropic has begun early-stage work on its own AI chip and held talks with Samsung Electronics as a potential manufacturing partner, following the lead of rival OpenAI in trying to get more control over the costly computing systems behind its models, according to three people with direct knowledge of the project.

If the Claude maker goes ahead with the chip, it would be a relative newcomer in developing its own AI server chips, compared to others. Google and Amazon Web Services have both been successful at developing their own chips over many years, while Meta Platforms and Microsoft have also developed their own chips. OpenAI tapped Broadcom to design its own chip in 2024 and last month unveiled the first product of that partnership, Jalapeño, an inference chip built to run large-language models more efficiently.

Anthropic is still at the stage of figuring out what it wants the processor to do, how powerful it should be and how it would fit into a server or clusters of servers, according to the three people. It has held discussions with multiple chip design firms but it has yet to move into detailed design, testing and manufacturing work.

Such processors are hard to design because engineers must balance speed, power consumption, memory, networking and cooling all at once, and manufacturing reliably at high volume is even harder.

The project underscores a broader shift where AI developers such as Anthropic are looking to control the infrastructure behind their models, including chips, cloud contracts, power and data centers. The biggest AI models run on vast clusters of processors, and at that scale even small efficiency gains can lower costs and free up scarce computing capacity. An in-house AI chip would add leverage as AI developers compete for processors, data-center space and electricity.

It’s possible Anthropic may not proceed with the chip, although the company has begun hiring for chip engineers. Earlier this month, it hired Clive Chan, an early member of OpenAI’s custom chip team.

In response to questions from The Information, Anthropic said Amazon Web Services’s Trainium chip, Google tensor processing units and Nvidia graphic processors will remain central to how the company scales its compute strategy. The company declined to share more about its roadmap. Samsung declined to comment.

The Takeaway

Tesla told staff last month it would impose a $200 per week limit for staff’s AI spending beginning July 6, according to an internal memo, a sign that even companies committed to using the technology to transform their operations and products are having to watch their costs.

Over the past few months, software engineers at Tesla were often consuming thousands of dollars’ worth of tokens each week, according to two people familiar with the usage. Workers will need a sign-off to spend above the new limit, the memo said, though the tally excludes beta versions of xAI products, the two people said.

The move follows an internal AI adoption push at Tesla. Over the past six months, leaders have worked to shift scattered employee usage to a companywide approach, with approved models and formal security policies, people who worked with the technology said, followed quickly by imposing the new guardrails on spending.

What’s going on inside Tesla mirrors the whiplash playing out across many businesses as companies like Meta Platforms, Uber and Walmart move from pushing employees to go all-in on using AI to limiting their spending. How quickly Tesla’s arc played out is striking, though, as it initially lagged behind some tech and other corporate giants in its steps to formalize workers’ AI usage and policies.

Tesla’s internal AI rollout is also particularly high stakes, since it’s playing out against the backdrop of the company’s broader vision of using AI to power its next generation of products. Musk has said Tesla’s future value depends on its ability to deploy AI at scale across its Robotaxi network and Optimus humanoid robot, not on selling cars. The company’s revenue has mostly stalled over the past two years.

But Tesla’s efforts to encourage workers to use AI have proceeded in fits and starts. The company launched a centralized platform for employees to access AI models last year, four people with knowledge of the platform said. The internal system, known as Bottle Rocket, provided access to models from OpenAI and Anthropic, as well as xAI and Cursor models, including unreleased versions. Before that, some employees used personal accounts, one person said.

Until recently, policies around internal usage were also piecemeal, with guidance on AI use determined at the vice president or director level rather than through any companywide policies, the four people said. Last year, Tesla’s former vice president of IT, Raj Jegannathan, was tasked with leading Tesla’s AI push and incorporating it into more departments, including the sales and service division, where there was a push to use AI agents for customer service, three people with knowledge of the efforts said. Several months before he left Tesla, Jegannathan lost some responsibilities and was no longer leading the sales team, one person said. After Jegannathan departed in February, Tony Tran began reporting to Musk as the director of IT, AI and cloud infrastructure.