A new Bitquery analysis reveals over 96,000 fake-address outputs clogging Bitcoin's unspent transaction database, raising concerns about node efficiency and the long-term impact of persistent unspendable outputs on network health
Bitcoin's unspent transaction output (UTXO) database is carrying a hidden burden: more than 96,000 outputs masquerading as spendable coins, but in reality, they're nothing but embedded text with no private key. This persistent bloat, uncovered by blockchain data firm Bitquery, is quietly inflating the state that every full node must track-an operational cost that doesn't disappear with time.
How Text Ends Up in Bitcoin's UTXO Set
Bitquery's scan of 965,135 Bitcoin blocks identified two main ways text is embedded on-chain: through OP_RETURN outputs and through fake-address outputs. OP_RETURN is a protocol-sanctioned method for attaching arbitrary data to a transaction, but these outputs are immediately pruned from the UTXO set after confirmation. In contrast, fake-address outputs embed readable text where an address hash would normally go, tricking the network into treating them as spendable coins. Since there's no corresponding private key, these outputs are effectively unspendable, yet they persist in the UTXO set indefinitely.
The difference is not academic. While OP_RETURN outputs expand the blockchain's historical record, they don't force every node to track them as live coins. Fake-address outputs, on the other hand, create a permanent drag on node resources, as each one must be indexed and checked for potential spending. Bitquery's census found 96,231 such outputs across 3,286 transactions, holding roughly 3.2 BTC that is, for all practical purposes, lost to the network.
Taunts, Fees, and Node Impact
One campaign in March 2026 targeted Bitcoin Core developer Luke Dashjr, embedding a two-sentence taunt in both OP_RETURN and fake-address outputs. Bitquery counted 13,062 OP_RETURN taunt transactions-each from a unique sending address-spanning 579 blocks and costing a total of 0.1014 BTC in fees. The fake-address outputs, however, are the ones that linger, continuing to occupy space in the UTXO set and increasing the workload for every node operator.
Bitcoin Core 30.0, released in October 2025, made it easier to relay larger OP_RETURN transactions by raising the default datacarriersize to 100,000 bytes and allowing multiple OP_RETURN outputs per transaction, up to an aggregate limit. This policy change did not alter consensus rules, but it did make it more practical for users to embed longer messages. Bitquery observed that messages exceeding the previous 80-byte ceiling were rare before v30, but surged into the thousands by February 2026. Still, the most persistent node-state burden comes from fake-address outputs, not OP_RETURN data.
Data Gaps and Broader Context
Bitquery's analysis excluded coinbase fields for 191,427 historical blocks and did not count witness inscriptions, focusing instead on four main channels for text insertion. The study's findings highlight a technical trade-off: OP_RETURN outputs expand the blockchain's archival footprint, but fake-address outputs create a lasting operational cost for every node. This distinction is not just theoretical-node operators must allocate more storage and processing power to maintain an ever-growing UTXO set, which can affect network decentralization and accessibility over time.
For comparison, other blockchain networks have faced similar challenges when protocol upgrades or user behaviors increase the state that nodes must track. As reported earlier, Solana's Agave 4.2 client hit its activation window, but mainnet upgrades affecting state management remain pending, leaving users and developers watching for concrete changes.
What the Numbers Show
According to Bitquery, the 96,231 fake-address outputs identified across Bitcoin's history are spread over 3,286 transactions and collectively hold about 3.2 BTC. In contrast, the 13,062 OP_RETURN taunt transactions paid a total of 0.1014 BTC in fees. The broader scan classified 4,412,782 OP_RETURN outputs as readable text, but only the fake-address outputs remain in the UTXO set, directly impacting node resource requirements. These figures are based on a scan of 965,135 blocks, with the most significant spike in large OP_RETURN messages occurring after the October 2025 release of Bitcoin Core 30.0.
Bitcoin's UTXO set is the backbone of its transaction validation process, ensuring that coins are not double-spent. Every full node must maintain a complete and up-to-date record of all unspent outputs, which means that persistent fake-address outputs increase the hardware and bandwidth requirements for running a node. Over time, this can make it harder for individuals to participate in network validation, potentially concentrating power among larger operators with more resources. The continued presence of unspendable outputs is not just a technical curiosity-it's a real cost that shapes who can run a node and how decentralized the network remains.
