Anthropic reshuffles leadership to back internal experimentation lab

When the co-founder of Instagram voluntarily relinquishes a C-level title at one of the hottest AI companies on the planet to become simply a 'member of the technical staff,' it's not a demotion. It's a signal that the company has found something it considers more important than corporate hierarchy.

Mike Krieger, who joined Anthropic two years ago as Director of Product, is stepping down from his post. But not from the company. He is transitioning to the role of co-technical lead of Anthropic Labs' internal incubator — a division that the company has decided to rapidly scale. Ben Mann, previously heading product engineering, will co-lead the team. Ami Vora, the current head of product, will take over the vacant position as head of the product division. Krieger will report directly to Anthropic's President Danielle Amodei — a fact that speaks to the strategic weight of his new appointment.

To understand the scale of this decision, context is needed. Anthropic Labs emerged in mid-2024 as a tiny initiative — literally two people working on experimental projects at the intersection of research and product. A year and a half later, company leadership decided that experiments deserve one of Silicon Valley's most experienced product leaders. Krieger is not simply a manager; he is the person who, alongside Kevin Systrom, built Instagram from zero to a billion users. His transition to the laboratory is not a quiet exile, but a concentration of the most valuable resource on a direction that Anthropic considers strategic.

The focus of Labs is 'experimental products.' Behind this vague definition lies a concrete logic. Anthropic has so far earned primarily through Claude's API and its consumer chat interface. But the market for large language models is rapidly commoditizing: OpenAI, Google, Meta, xAI, and dozens of startups offer models comparable in quality, and competing solely at the level of underlying technology is becoming increasingly difficult. The internal incubator is an attempt to find new product forms for AI, to move beyond the familiar chatbot and API access. In essence, Anthropic wants to understand what next-generation AI applications will look like and build them before competitors do.

This strategy is not unique, but its execution is atypical. Google has long experimented with AI products through DeepMind and internal labs, but it does so within a gigantic corporate structure where innovation often gets buried in bureaucracy. OpenAI, by contrast, launches new products aggressively, but from a single product center. Anthropic chooses a third path: it isolates an experimental team in a separate structure with direct access to leadership, but within a relatively compact organization. This preserves startup speed while giving experimenters freedom from the pressure of current product metrics.

There is also a personnel aspect that cannot be ignored. Shuffles at the C-suite level always raise questions about internal politics. However, the fact that Krieger remains at the company and gains direct access to Danielle Amodei suggests a well-thought-out strategic maneuver rather than conflict. Ami Vora, who is assuming product leadership, is an experienced leader, and delegating operational control of mature products to her looks like a logical division of labor: one person is responsible for what brings money today, another for what will bring money tomorrow.

For the industry, Anthropic's move is another confirmation that the race for large language models is entering a new phase. Pure model power ceases to be the chief differentiator. Victory will go to those who find the right product packaging for AI — interfaces, use cases, and integrations that users haven't even imagined yet. Creating a dedicated lab with top-tier leadership is a bet that AI's future will be determined not by benchmarks, but by product breakthroughs.

Anthropic Labs' specific experiments remain undisclosed. But the very fact that senior leadership is being reorganized for this direction speaks louder than any press release. Anthropic clearly saw something on the horizon — and decided it needed someone who has already once changed how hundreds of millions of people interact with technology.