Broadcom CEO Stands Firm on $230B AI Revenue Targets Despite Anthropic's Slowdown Push

Broadcom Reaffirms AI Revenue Targets Amid Industry Slowdown Debate

Broadcom CEO Hock Tan stood firm on the chipmaker's ambitious AI revenue targets despite growing market concerns triggered by Anthropic CEO Dario Amodei's weekend essay advocating for a moderation in AI model development1

1

. Speaking on CNBC's Mad Money on September 14, Tan dismissed speculation that the slowdown push would impact Broadcom's fiscal 2027 and 2028 AI semiconductor revenue forecasts, stating demand for AI compute infrastructure remains "very strong" and "very durable"1

1

. The chipmaker maintains its projection of $115 billion in AI semiconductor revenue for fiscal 2027, doubling to $230 billion in fiscal 20281

1

.

Market Reaction and Semiconductor Stocks Under Pressure

Investors reacted sharply to Amodei's essay, which received public support from OpenAI CEO Sam Altman and Elon Musk2

2

. Broadcom shares fell 4.8% on Monday, while the broader iShares Semiconductor ETF declined 5.6%2

2

. The selloff extended across AI infrastructure providers as investors reconsidered their expectations for compute demand1

1

. Broadcom's sensitivity to Amodei's commentary stems from Anthropic's position as a critical custom chip customer. Tan revealed that Anthropic is on track to become Broadcom's largest custom chip customer in 2027 and sustain that designation in 2028, with visibility to 10 gigawatts of deployment by 20283

3

.

Inference Demand Drives Broadcom's Confidence in AI Growth

Hock Tan made a critical distinction between training new frontier AI models and inference, the day-to-day usage of deployed models1

1

. While acknowledging uncertainty around training demand, Tan expressed particular optimism about inference: "When you want to productize inference, I see it continuing to be very, very strong"1

1

. This distinction matters because even if the pace of AI model development slows, existing models still require massive compute infrastructure for inference operations3

3

. Every AI query and enterprise AI agent deployment demands ongoing compute resources, creating durable demand independent of training slowdowns3

3

.

Broadcom's Position on AI Safeguards and Development Pace

Tan acknowledged the need for AI safeguards while pushing back against existential risk concerns at current capability levels3

3

. "Like any tool, it's important to put governance and safeguards on how we use the tool," Tan told Cramer, but emphasized that AI "is not a live animal that will run wild by itself"1

1

. Drawing a historical parallel, Tan compared generative AI's potential impact to the Industrial Revolution, suggesting the technology will "create huge value" and help humanity "reach a better level of living"1

1

. This framing positions AI as a transformative but manageable tool rather than an uncontrollable force3

3

.

Capital Commitments Signal Continued AI Infrastructure Investment

Despite the slowdown debate, capital continues flowing into AI infrastructure at unprecedented levels. Tan helped establish a $35 billion AI special-purpose vehicle platform with Apollo and Blackstone to support Anthropic's one-gigawatt deployment3

3

. This massive financial commitment underscores investor confidence in long-term AI infrastructure demand, even as industry leaders debate the appropriate pace of model advancement. Broadcom's AI revenue consists of both custom AI accelerators and networking chips used in AI systems, positioning the company across multiple layers of the AI infrastructure stack1

1

. Google has historically been considered Broadcom's biggest custom chip customer through co-designing tensor processing units, but Anthropic's trajectory suggests a shifting landscape in custom chip partnerships1

1

.

What This Means for AI Infrastructure Investors

The tension between Amodei's caution and Tan's confidence reveals a fundamental question facing AI infrastructure investors: does moderating frontier model development reduce compute demand or simply shift it toward inference and deployment? Broadcom's fiscal 2026 third-quarter results showed 86% year-over-year revenue growth, providing factual support for Tan's optimistic projections3

3

. However, Broadcom shares remain down 1.44% year-to-date and 5.97% over the past year, reflecting an expectations gap between the company's growth trajectory and market sentiment3

3

. Watch for continued customer deployment announcements and inference workload growth as key indicators of whether Tan's AI chip revenue target holds despite the broader slowdown in AI development discussion.