Anthropic is continuing to expand the role of its Claude AI platform as the company gains momentum among enterprise customers, pursues a funding round that could value the company at about $950 billion, and increases its focus on coding tools, automation, and proactive AI systems.
The company has emerged as one of the main competitors to OpenAI during a period when much of the technology industry remains centered on AI model development. OpenAI was valued at $854 billion during its March funding round, while Anthropic is reportedly seeking a significantly higher valuation in its next financing effort.
A recent report also found that Anthropic has recently surpassed OpenAI among business customers, with the company quadrupling its market share since May 2025. Businesses are increasingly showing preference for Anthropic’s Claude models over ChatGPT in workplace settings.
Claude Product Team Expands AI Coding Strategy
Cat Wu, Anthropic’s head of product for Claude Code and Cowork, has played a central role in the company’s recent product development efforts since joining in August 2024. Wu has helped oversee Claude’s transition from a general informational chatbot into a broader coding and productivity platform.
Wu works closely with Boris Cherny, a member of Anthropic’s technical staff and creator of Claude Code. The pair have been described internally as Anthropic’s “Batman and Robin” due to their involvement in Claude’s development.
During an interview at Anthropic’s second annual Code with Claude conference in San Francisco, Wu discussed the company’s approach to product strategy and future AI deployments.
Wu said Anthropic prioritizes maintaining pace with rapid AI model improvements instead of directly reacting to competitors.
“The main thing that we design for is staying on the exponential,” Wu said. “We don’t think about competitors. I think if you do think about competitors, you end up being, like, perpetually two weeks, or like, a month behind how fast you can execute.”
She added that Anthropic expects AI models to continue improving steadily and said the company wants to distribute those improvements broadly while still handling deployment carefully.
Anthropic Discusses Restricted Release Of Cybersecurity Model
Wu referenced Anthropic’s handling of Glasswing, an initiative launched in April that gave a limited group of organizations access to the company’s cybersecurity-focused AI model, Mythos.
The Glasswing consortium included companies such as Amazon, Apple, CrowdStrike, and Microsoft. Unlike many of Anthropic’s public-facing models, Mythos was not released broadly.
Anthropic said the model, which is designed to scan software codebases for vulnerabilities, could potentially be weaponized by malicious actors if distributed widely.
Wu said the company wants AI intelligence to benefit as many users as possible, but added that deployment decisions must also account for safety concerns.
Anthropic Sees Human Workers Managing AI Agents
Wu also discussed Anthropic’s view of workplace automation and AI agents. In response to questions about whether agents could eventually outperform the humans supervising them, Wu argued that managers would still need strong expertise in their own fields.
“I think it is extremely hard to manage agents if you can’t do the job yourself,” Wu said.
According to Wu, managing AI agents resembles managing human employees because users still need to understand why mistakes happen, whether instructions were unclear, or whether requests lacked sufficient detail.
When asked whether AI agents could eventually reduce hiring needs, including entry-level positions, Wu said the company instead views AI systems as tools that can remove repetitive work from existing jobs.
Wu cited email management and other routine tasks as examples of work that AI agents could automate, allowing users to spend more time on projects they want to pursue.
Anthropic Plans More Proactive AI Systems
Looking ahead, Wu said Anthropic’s next major focus involves making Claude more proactive.
She described last year’s AI development cycle as centered on synchronous interactions, where users actively request assistance in real time. According to Wu, users are now increasingly shifting toward routines and automated workflows, including customer support automation.
Wu said the next stage is for Claude to understand an individual user’s work patterns and independently establish automations based on those activities.
Featured image credits: Heute.at
For more stories like it, click the +Follow button at the top of this page to follow us.