Everyone at the HumanX conference was talking about Claude

This week, thousands of technology professionals gathered at San Francisco’s Moscone Center for the HumanX AI conference, where the conversation centered on the transformative impact of agentic AI on business. These agents, which automate complex business and coding tasks, are increasingly being integrated across various industries, primarily via enterprise-level and consumer-facing chatbots.

Naturally, I wanted to know which chatbot was the most popular, and I kept hearing one name over and over again: Claude.

Anthropic received frequent mentions across various panels throughout the week and was a recurring subject of conversation among the vendors I engaged with on the convention floor. Noticeably absent from these discussions, however, was ChatGPT. One vendor, in particular, noted that his team now relies heavily on Claude, suggesting that ChatGPT and OpenAI have declined in quality—or, in the parlance of the internet, have “fallen off.”

This sentiment is no longer an isolated opinion. It remains unclear how OpenAI can reverse the growing perception that it has lost its momentum, particularly given its recent $122 billion funding round and the anticipation surrounding its upcoming IPO. The company appears increasingly uncertain about its next strategic move.

Part of the issue may stem from a perception that the company lacks strategic focus. Last month, OpenAI discontinued several long-running side projects—including its Sora video generator and a controversial plan for a “sexy” version of ChatGPT—to prioritize its business and coding services. Simultaneously, the company has faced a wave of negative publicity, driven by a New Yorker investigation that questioned CEO Sam Altman’s trustworthiness, as well as criticism regarding its ties to the Trump administration and the introduction of advertisements into ChatGPT.

When Alex Heath asked about the profile in the New Yorker at the HumanX AI conference, Brett Taylor, chairman of the board of OpenAI and CEO of Sierra, strongly supported Altman. “I think Sam is one of the most well-known leaders and executives in the world,” Taylor said.He also acknowledged that celebrities naturally attract harsh critics. “I think Sam is extraordinary. I think he’s an extraordinary leader in AI, and having worked with him, I have complete confidence in his character.”

These ongoing controversies and policy shifts can give the impression that OpenAI is acting reactively rather than strategically, as if it is merely responding to external events instead of driving the agenda. However, in terms of prominence and revenue, OpenAI and Anthropic remain closely matched. A recent Wall Street Journal analysis labeled them the “fastest-growing businesses in the history of tech,” and data suggests Anthropic is rapidly gaining ground with business users. In this light, perhaps OpenAI’s perceived decline simply marks its transition from an undisputed champion to a major player in a more competitive, and therefore more normal, market landscape.

OpenAI remains clearly committed to maintaining its industry dominance. This week, the company unveiled a new $100 monthly subscription tier for ChatGPT, featuring significantly expanded access to its Codex coding tool. This strategy appears intended to boost adoption of the service while actively competing to win over users currently utilizing Claude Code.

In a conversation with Bloomberg’s Rachel Metz at the HumanX conference, Srinivas Narayanan, OpenAI’s CTO of B2B applications, remarked on the rapid pace at which the technological landscape is evolving.

“We are currently in an extraordinary technological era where we expect new developments every month, if not every day,” Narayanan remarked. Citing agentic coding as a prime example, he added, “We anticipated that AI would transform software engineering, and while people have utilized assistive coding tools for the past year, the entire field has undergone a significant transformation in just the last few months.”

Agentic capabilities have become a central focus for the tech community, in part because other AI applications—such as creative tools—have yet to fully deliver on their potential. Nevertheless, the scale at which companies have already delegated workloads to these automated agents is remarkable, particularly given the short timeframe in which this transition occurred, as noted by Narayanan. In such a volatile landscape, the future remains highly unpredictable.