Scan reveals 96,000 fake-address outputs clogging Bitcoin’s unspent database

Blockchain information firm Bitquery printed a census on Sept. 2 of 4 methods textual content seems in Bitcoin, scanning 965,135 blocks and revealing that the identical phrases can impose very completely different prices on nodes.

Its study counted 13,062 OP_RETURN taunt transactions that added no spendable state, towards 96,231 fake-address textual content outputs that stay within the unspent transaction output set.

The March 21-26, 2026, marketing campaign named Bitcoin Core developer Luke Dashjr in the identical two-sentence taunt. Bitquery counted 13,062 transactions from 13,062 sending addresses throughout 579 blocks, with 0.1014 BTC paid in mixture charges. The chain data the textual content, timing, inputs, and costs, however it doesn’t establish the payer, creator, or motive.

Across the broader scan, Bitquery labeled 4,412,782 OP_RETURN outputs as readable textual content. Their node price differs from fake-address textual content as a result of Bitcoin treats the 2 output sorts otherwise after affirmation.

Nodes don’t have to retain an OP_RETURN output within the UTXO database used to examine whether or not cash could be spent, though its information stays in Bitcoin’s block historical past. The Bitcoin developer guide describes this null-data development as preferable to strategies that create outputs which look spendable and burden the UTXO set.

Fake-address textual content locations readable bytes the place an deal with hash would usually go. A node sees an apparently spendable output, whereas the embedded textual content has no identified corresponding non-public key.

Bitquery labeled 96,231 such outputs throughout 3,286 transactions and estimated that they maintain about 3.2 BTC that’s successfully unspendable. Those outputs persist as present node state except they’re spent.

Bitcoin Core 30.0, launched in October 2025, made bigger OP_RETURN transactions simpler to relay underneath its default coverage. The release notes raised the default -datacarriersize setting to 100,000 and allowed a number of OP_RETURN outputs underneath an mixture restrict for relay and mining.

These standardness decisions didn’t change the consensus guidelines that decide whether or not a block is legitimate.

Bitquery discovered that messages longer than the previous 80-byte payload ceiling appeared just a few dozen occasions in most months earlier than v30 and rose into the hundreds by February 2026. Easier default propagation gives one mechanism for that rise, however the observational sequence doesn’t present that v30 alone triggered it.

The March taunt disclosed textual content is 62 ASCII bytes, which already match beneath the previous ceiling. The marketing campaign measures the low price of repeating OP_RETURN textual content after v30.

Bitquery additionally lacked coinbase fields for 191,427 historic blocks and didn’t depend witness inscriptions. The examine due to this fact compares the state penalties of 4 measured channels quite than tallying each piece of textual content embedded in Bitcoin.

Within that scope, OP_RETURN expands block historical past with out bloating the UTXO set, whereas fake-address textual content creates the lasting node-state burden.

The put up Scan reveals 96,000 fake-address outputs clogging Bitcoin’s unspent database appeared first on CryptoSlate.