The Ethereum Foundation lost two more researchers on Monday when Carl Beekhuizen and Julian Ma announced their departures on X, pushing the number of senior exits this year to at least nine - and triggering a public debate about whether the organization that has stewarded Ethereum's development for a decade can still execute the protocol's roadmap.
Beekhuizen, who spent seven years at the EF, contributed to the early design of the Beacon Chain and organized the KZG ceremony underpinning Ethereum's data availability infrastructure. "After 7 incredible years, I've decided that Friday May 29th will be my last day at the Ethereum Foundation," he wrote on X. Ma, a researcher whose work shifted toward product and growth over time, said the EF "is an amazing place but not right for my next steps," per his X post. Senior solutions architect Pablo Voorvaart followed the next day, Tuesday May 20.
The three-in-two-days cluster extends a wave that started in February. Co-Executive Director Tomasz Stańczak announced on February 13 that he would step down at the end of the month, less than a year after the EF introduced its dual-leadership structure. Dankrad Feist departed to join stablecoin L1 project Tempo. Security architect Josh Stark also left. Five of the nine departures have come in May alone.
The exits follow the EF's publication on March 13 of the CROPS Mandate - a 38-page document the Foundation described as "part constitution, part manifesto, part guide." The mandate codified four non-negotiable principles: censorship resistance, open source, private, and secure. Several contributors were reportedly asked to align formally with the document; community observers have linked the friction it created to some of the subsequent departures.
The structural question underneath the personnel numbers is whether the exits represent a deliberate redesign or an uncontrolled unraveling. The case for planned transition comes largely from Ryan Berckmans, who pushed back on X against alarm in the community: the time for new blood "is here," he wrote, and Vitalik Buterin's 2025 reorganization - which pushed execution outward to independent client teams and recast the EF as a focused research and grants hub - is proceeding as designed. On that read, experienced researchers moving to client teams, L2 projects, and competing chains is the model working, not failing.
But prominent voices are less sanguine. Laura Shin and Bankless co-founder David Hoffman have both raised the concern that the EF's ideological rigidity - its explicit refusal to compete on commercial terms or optimize for institutional throughput - is bleeding talent to projects with clearer commercial mandates. Dankrad Feist, now outside the EF, has gone further: he proposed on X creating an independent $1 billion ETH-funded organization with a board "who want ETH to go up" and a leader "who is competent and wants to fight" - a direct critique of what he sees as the EF's structural misalignment with ETH's competitive position.
The practical risk centers on the Glamsterdam upgrade and post-quantum security work - both named by Berckmans as areas where the EF remains "intensely focused." Banteg captured the community's unease plainly: "situation: all three ef protocol leads have left," he posted on X alongside a strikethrough of the EF org chart.
Whether that concern is justified depends on how much of Ethereum's roadmap execution has already migrated to independent client teams - and whether those teams can absorb the institutional knowledge that left with Beekhuizen, Feist, and their cohort. The EF has not made a public statement on the departures.
Editorial note for reviewer: The task brief stated May 22 as the event date, but primary sources (Carl Beek's X post, Unchained's article dated May 20 stating announcements occurred "Monday") confirm the announcements were May 19, 2026. The piece uses May 19. The brief-stated count of eight was accurate at the time of Beek/Ma's departure; with Voorvaart the verified count is now nine. The 25% headcount figure mentioned in the prior editorial round has no traceable primary source connecting it to the EF specifically - that figure applies to the Algorand Foundation per public records. It has been omitted.