Want to know the most under-reported story in crypto? It's not some altcoin pumping. It's not a new L2. It's the quiet, almost invisible shift in who actually calls the shots on Ethereum.
I saw it first-hand last month at a small validator meetup in Prague, tucked away in a beer cellar near the Vltava. The conversation wasn't about EIP-4844 or Danksharding. It was about who funds the client teams, who runs the RPC endpoints, and who gets to decide what 'final' means. The Ethereum Foundation isn't dead. But its monopoly on governance? That's bleeding out in slow motion. And most people are still looking at the price chart.
The Old Guard
Since 2015, the Ethereum Foundation (EF) has been the de facto steward. They funded research, coordinated core dev calls, held the treasury, and set the narrative. Vitalik’s voice carried weight because he was the architect. The EF’s Swiss structure gave it legal cover. Its quarterly reports showed where the money went. For years, if the EF wanted an EIP to pass, it passed. If it wanted a research topic, it got funded. That was the single-node model.
But the Merge changed everything. Proof-of-stake introduced a new class of power-holders: validators. Not just anyone—large staking pools like Lido, exchanges like Coinbase, and solo stakers. Suddenly, the people who secure the network also have a say in its future. Combine that with the rise of independent client teams (Geth, Nethermind, Besu, Erigon) and infrastructure giants (Infura, Alchemy), and you have a multi-polar world. The EF is now just one node among many.
The New Power Map
Let me break down the new power map using data I’ve tracked since my days at the 2024 ETF flow desk. First, client diversity. Geth still dominates with about 80% of execution clients. That’s a single point of failure not just for tech but for governance. If the Geth team decides to take a stand on an EIP—say, one that reduces their maintenance burden—they hold immense leverage. Second, staking pools. Lido holds over 30% of all staked ETH, and its stETH tokens are the liquidity backbone of DeFi. Through its DAO—which has notoriously low voter turnout—it can influence protocol decisions. Third, infrastructure providers like Infura and Alchemy act as gatekeepers. They serve the majority of dApp traffic. If they choose not to support a contentious hard fork, they can effectively veto it. Fourth, the EF itself still holds a multi-billion dollar treasury and funds key research, but its influence is waning relative to these actors.
I’ve seen this movie before. In 2017, during the Ethereum Classic fork, the power struggle was between miners and developers. Back then, I wrote a 500-word breakdown within 12 minutes of the fork activation—not because I had deep code knowledge, but because I understood sentiment. The same dynamic plays out now, just with different players. The 'silent' nature of this shift means that when a decision needs to be made—say, on a controversial state expiry proposal or a change to the fee market—the EF can no longer just announce it. It has to negotiate with Lido, with Geth, with Infura.
Social capital outpaced code in the ape arcade. Now governance capital is outpacing technical merit in the L1 theater.
Let’s talk numbers. According to recent community surveys, the EF’s share of core development contributions has dropped from over 70% in 2020 to around 40% now. The rest comes from independent teams and commercial entities like Nethermind (which is for-profit) and Besu (backed by Hyperledger). Meanwhile, Lido’s staking share has plateaued at around 30%, and the top 10 staking entities control over 60% of validators. That’s not a network of a million nodes—it’s an oligopoly of a dozen.
Speed is the only metric that survived the crash. But here, the crash might be a governance crisis.
The EF’s slow and deliberate approach has been replaced by a more fragmented process. That could lead to faster decisions? Unlikely. More likely, it leads to gridlock. Just look at the ongoing debates around EIP-7251 (max effective balance increase) and the push for a 'runway' for client teams. Delays are already happening. The core devs are spending more time on process than on innovation.
The Contrarian Angle: The New Oligarchy
Here’s what nobody’s talking about: this 'silent coup' might actually be bad for Ethereum’s security. Not because the new rulers are malicious, but because they are not accountable. The EF had a formal structure, transparency reports, and a reputation to protect. Lido’s DAO is a messy multi-sig with low voter participation. Geth’s maintainers are a small team with no formal governance. Infura is a for-profit company owned by ConsenSys. If these entities collude—even unintentionally—they can centralize power without any checks. The outcome could be a 'governance attack' where a small group pushes through a change that extracts value from regular holders. And there’s no way to stop it because there’s no voting. It’s all off-chain influence.
Reading the room while the order book burns—except the room is a private Signal group, not a public forum.
Furthermore, this shift weakens the 'non-security' argument. The SEC has said that a network is sufficiently decentralized when no single entity can control it. But if a small group of nodes can coordinate, the network is effectively controlled by a few. That could re-introduce securities risk. The EF was a single entity, but it had a clear mission. Now we have multiple conflicting interests. Who do regulators call? That uncertainty is a risk premium that ETH holders aren’t pricing in.
My Take: What to Watch
I lived through the 2022 FTX collapse and saw how quickly narratives can flip. The same applies here. The next six months will be critical. Watch for three signals:
- A shift in EF budget allocations. If the Foundation announces it will reduce funding for core protocol R&D and focus on community grants, that’s a clear sign they are ceding control to other nodes.
- A contested EIP that stalls. If a proposal that has strong support from one group (say, large stakers) but opposition from client teams gets stuck, the fault lines become visible.
- Validator concentration metrics. If Lido’s share climbs above 33% or if a single client like Geth approaches 90%, the oligarchy risk becomes acute.
The sprint doesn’t end when the block confirms—it ends when the next governance battle begins. And that battle is already being fought, just not on the timeline.
So stay frosty. Keep your stETH diversified. And next time you see a core dev tweet about an EIP, remember: they’re not just talking about code. They’re signaling who holds the pen.