Anthropic, the AI company preparing for a public listing, has begun an official collaboration with the Budapest-based, Hungarian–American startup Zocks. According to Zocks’ statement to Telex, the Budapest startup’s virtual assistant will be accessible within the Claude version designed for financial advisors after Anthropic presented the development on Monday in California.
What Zocks brings to Claude
Founded in 2022, Zocks builds privacy-focused AI assistants primarily for financial firms, advisors and asset managers. The company’s assistant automates previously manual tasks, takes notes, converts client conversations, emails and documents into structured data and sends reminders, which the startup says makes daily work significantly faster and more efficient.
Under the announced integration, Claude will receive full contextual information through Zocks: real client data, plans shared by clients, priorities and concerns. As a result, advisors will not need to craft highly detailed prompts, and Claude will be able to execute tasks more effectively using the client-specific context.
Zocks’ market position and team
Zocks ranked fifth on Forbes’ 2025 list of “Hungary’s hottest startups.” Forbes noted at the time that Zocks’ product can be six to seven times more expensive than typical competitors, which led the company to target the financial sector where customers tend to be less price-sensitive. The founders said they are aiming primarily at the US market while building their development team in their Budapest office. Zocks maintains offices in Budapest and San Francisco; its founders are Ratku Ákos and Mark Gilbert.
Anthropic’s situation and the surrounding debate
Anthropic is one of the world’s fastest-growing and most closely watched tech companies: in May the firm’s estimated valuation was 965 billion dollars. The company is preparing for an IPO that may take place in the coming weeks.
However, tensions flared in early September after a former Anthropic researcher posted on X that the current competitive race in AI development could pose serious risks to humanity. A few days later, Anthropic published a threat-analysis report detailing how various actors had used Claude for a range of purposes, from weapons development and cyber operations to surveillance and fraud.
Dario Amodei, Anthropic’s CEO, subsequently urged companies in the AI sector to slow the pace of capability development given growing concerns about misuse. In an essay shared on X he wrote: “Progress will still feel rapid, and we should use the time wisely.”
The Economist commented that slowing development might not necessarily harm Anthropic’s business, since current models are already sufficiently capable for many enterprise clients who account for a large share of the company’s revenue. In other words, further rapid model improvements may not be required for the company to continue growing revenue.
Why this matters
The partnership connects a specialized, finance-focused assistant developed by a Hungarian startup with a major international language model product, highlighting both the global reach of local developers and an industry trend toward integrating large models with sector-specific tools. At the same time, the debate around risks and misuse at Anthropic underscores ongoing concerns about safety and governance as the market expands quickly.
