Broadcom CEO affirms AI revenue targets unchanged despite Anthropic CEO's recent public call for slowing AI development.

Broadcom CEO Hock Tan dismissed concerns that a potential slowdown in frontier AI model development could impact the chipmaker’s business, reaffirming the company’s long-term revenue targets. His comments followed investor unease sparked by Anthropic CEO Dario Amodei’s weekend essay, which advocated for moderating the pace of AI model development. Amodei’s proposal, endorsed by OpenAI CEO Sam Altman and Elon Musk, prompted investors in AI infrastructure providers to reassess their compute demand expectations during Monday’s trading session. Consequently, Broadcom shares declined 4.8%, while the iShares Semiconductor ETF dropped 5.6%. Broader data-center component stocks also faced significant selling pressure.

When asked on CNBC’s “Mad Money” whether the slowdown debate had altered Broadcom’s fiscal 2027 and 2028 AI semiconductor forecasts, Tan responded, “No, not in the least.” He added, “We see the demand for compute infrastructure, for AI development or AI frontier models, and inference for the products that they feed to the world, as continuing to be very strong and, I believe, very durable.” During Broadcom’s fiscal 2026 third-quarter earnings call on Sept. 2, Tan projected AI semiconductor revenue of $115 billion for fiscal 2027, followed by a doubling to $230 billion in fiscal 2028. The revised 2028 target, which exceeded prior expectations, stood out as a key positive in the earnings report. Broadcom’s AI revenue encompasses both custom AI accelerators and networking chips deployed within AI systems. Tan noted that Anthropic is positioned to become Broadcom’s largest custom chip customer in 2027 and maintain that status through 2028—a projection that underscores why Broadcom investors remain highly sensitive to Amodei’s remarks. Historically, Google has held the top spot as Broadcom’s largest custom customer, having co-designed the company’s tensor processing units.

Tan expressed particular optimism regarding demand for inference—the daily operational use of trained AI models. “I don’t know about training, but when you want to productize inference, I see it continuing to be very, very strong,” he said. Amodei’s essay intensified industry debate over whether AI development is advancing too rapidly, outlining a three-step framework to decelerate progress without compromising commercial advantage or U.S. leadership in AI. While Tan acknowledged the necessity of governance, he remained less concerned about the technology’s trajectory. “Like any tool, it’s important to put governances, safeguards on how we use the tool,” he told Cramer, adding, “It’s not a live animal that will run wild by itself.” Instead, Tan highlighted AI’s capacity to drive productivity gains. “AI, generative AI, the creation of those frontier models … will create huge value,” he stated, comparing the current wave of innovation to the 18th-century Industrial Revolution that originated in England. “It is still at the end of the day a tool that will make our society, humanity, reach a better level of living,” he concluded.