Ira Bodnar, founder of San Francisco-based startup Ryze, has said that a recent update to Anthropic's Claude AI effectively wiped out her company's core business "overnight."
Ryze had built a tool designed to help businesses manage and optimise advertising campaigns across Google and Meta platforms. Within just a few months of launch, the startup had secured several hundred paying customers and was closing around 70% of its sales deals in the competitive ad-tech market.
But that momentum came to a sudden halt.
According to Bodnar, new features rolled out by Anthropic's Claude AI, along with similar tools introduced by another AI firm, Manus, began offering capabilities that directly overlapped with Ryze's product. After the Claude update, Ryze's deal close rate reportedly dropped sharply from 70% to just 20%.
"It changed the landscape almost instantly," Bodnar suggested, describing how rapidly evolving foundation models can outpace startups building on top of them.
Rather than shutting down, Ryze is now pivoting. The company has shifted its focus to building complex AI workflows tailored for large advertising agencies that manage hundreds of client accounts with lean teams. Bodnar remains optimistic, arguing that while AI makes product development easier and faster, competitive advantage now lies elsewhere.
"Human taste and distribution strategy are becoming the real moats," she has said, suggesting that understanding how to position and scale products matters more than simply building them.
Bodnar also raised broader concerns about the future of online content. She predicted that social media platforms could soon be dominated by artificial activity, estimating that as much as 98% of videos on platforms like TikTok and Instagram may eventually be AI-generated and algorithmically amplified.
In that world, she believes success will depend less on originality and more on mastering platform algorithms and scaling content output efficiently.
Claude AI at the Centre of Policy Debate
Claude, developed by Anthropic, is positioned as a "helpful and harmless" AI assistant. It is known for its contextual understanding and ability to generate human-like responses, assisting users with writing, planning, coding and conversational tasks.
However, the company is currently facing mounting pressure from the US government over the military use of its AI systems.
Reports indicate that US Defence Secretary Pete Hegseth recently summoned Anthropic CEO Dario Amodei to discuss lifting restrictions on the military applications of Claude.
The Pentagon is reportedly seeking broader, potentially unrestricted access, including for uses that Anthropic considers red lines, such as fully autonomous weapons systems and mass surveillance of Americans.
Anthropic has signalled it has no intention of easing its safeguards. Amodei has reiterated the company's commitment to safety-focused AI development, stating that certain uses remain off-limits.
Hegseth has warned of "significant consequences" if Anthropic refuses to comply. According to reports, the Pentagon could designate Anthropic as a supply chain risk, a label typically applied to foreign adversaries, or invoke the Defence Production Act (DPA). The DPA allows the US government to compel companies to prioritise or expand production for national defence purposes.
The stakes are high. Last year, Anthropic signed a $200 million contract with the US government and is currently the only major AI provider cleared to operate on classified military networks.
Cutting ties could complicate Pentagon operations, though the Department of Defence has begun signing contracts with other AI firms, including Elon Musk's xAI.
A Rapidly Shifting AI Landscape
The twin developments, startups disrupted by foundation model updates and governments pushing for deeper AI integration into defence, highlight the speed and scale of the AI industry's evolution.
For founders like Bodnar, the lesson is stark: when building on top of powerful AI platforms, today's breakthrough feature can become tomorrow's built-in default.
And for companies like Anthropic, the challenge is balancing commercial growth, national security demands and the ethical boundaries they have publicly committed to uphold.
