Base, the Coinbase-backed Ethereum layer-2 network, activated its Cobalt upgrade on September 30, and while headline coverage focused on the fork itself, the more consequential changes sit inside the token standard that Cobalt supercharges. For issuers of B20 tokens, the upgrade adds explicit issuer powers over balances, and that makes the identity of an issuer more important than ever.
The upgrade went live at 18:00 UTC on September 30, one week after the Sepolia testnet activation, with Base’s status page marking the maintenance window complete at 20:00 UTC. Beyond the asset functions, Cobalt introduces validity transactions that can be set aside until specified on-chain conditions are met, a registry for scheduling future upgrades in monitoring mode, and an on-chain method for registering trusted-execution-environment prover signers. Node operators were required to run at least version 1.4.2 for mainnet.
What issuers actually gained
The asset-side additions are the heart of the story. B20 issuers gained two composite policy types, Union and Intersect, which let them combine existing transfer rules into more sophisticated logic. A scheduled multiplier function can change displayed token balances at a future time without each holder signing a transaction. And a new seizure operation can move a holder balance under issuer authority when the relevant policy permits it.
None of this makes a tokenized asset a share in the company it tracks. What it does is make the rules enforced by the token more explicit, and the consequences of issuer decisions more visible. A tokenized equity product can be represented as a balance on Base while rights to the underlying security sit with a broker, custodian, or contractual issuer. The token standard cannot force a transfer agent to recognize the wallet holder as a shareholder. That link comes from product documents and the entities responsible for backing, redemption, and corporate actions.
Seizure with a memo, not a verdict
Cobalt introduces seizeWithMemo, an issuer-authorized operation that moves tokens from a holder in one administrative step, superseding an earlier burn-and-block flow. The memo can leave a reason marker in the on-chain record. What it cannot do is prove the reason was legally sufficient. A smart contract can verify that the calling account has authority and that configured exemptions apply. It cannot decide whether a court order was valid, whether the issuer matched the correct defendant, or whether a customer’s complaint should succeed.
The issuer operations guide describes legitimate uses: a sanctions order, a mistaken issuance, recovery under contractual terms, or a corporate action. Each is a different justification, and holders should be able to find the administrator identity, the policy, the event, and a dispute process in the product’s legal documents. If an issuer only says tokenization is transparent, the obvious question is transparent about what. The transfer may be visible while the underlying decision remains opaque.
There is also a subtle integration detail that auditors should note. The exemption scope changed names from SEIZE_HOLDER_POLICY to SEIZE_EXEMPT_POLICY, with a different selector. Code that hardcodes the old scope can fail to read or set the new one, even though older Beryl-era selectors otherwise continue to work.
What did not ship
Two ideas that appeared in earlier Cobalt discussions are absent from the deployed scope, and conflating the roadmap with shipped code would be a mistake for institutions evaluating the standard. Payment of network fees in B20 tokens was removed from the fork’s list on September 29. Faster canonical 200-millisecond blocks belong to a proposed later upgrade known as Denim, not this activation. Native account abstraction has no scheduled mainnet gate here.
The node version floor is equally practical. Version 1.4.1 included the activation timestamp but missed changes to validity transaction RPC forwarding, and version 1.4.0 does not contain the mainnet activation at all. A node that follows the fork without forwarding the new transaction type correctly can present a partial view of what users believe is a uniform network.
Three layers of control, three owners of risk
The cleanest way to understand Cobalt is as a control split across three layers. The chain decides whether a transaction conforms to the deployed rules. The issuer decides which permitted administrative call to send. And the real-world asset provider is responsible for whether the token matches an enforceable claim. Cobalt changes the first two layers. It does not resolve the third. The same address may trade a token on-chain and still fail an off-chain eligibility test at redemption.
It is also worth stressing that the mere existence of the seize function does not grant every issuer power over every token. The B20 precompile reference is explicit: a token whose issuer has not configured the applicable policy slot has no seizure capability at all. Two assets using the same standard can carry sharply different holder rights, which is why token-level inspection matters more than standard-level headlines.
The strongest issuer case for these tools is straightforward: regulated assets need procedures for error correction and responses to legal orders, and the chain provides an evidence trail that conventional account corrections rarely expose publicly. If an issuer moves 100 tokens from one wallet to another, observers can count the tokens and identify the transaction. What they cannot infer is a matching entry in the issuer’s off-chain shareholder register without reconciliation.
A standard growing up in public
Cobalt is Base’s third mainnet upgrade, arriving after the Beryl upgrade introduced the B20 token format for stablecoins and tokenized assets. With Coinbase’s tokenized-stock push already live on the network, the audience for these controls is no longer hypothetical. The institutions now evaluating Base are asking compliance questions with legal departments attached, and Cobalt’s answer is that issuer powers can be precise, auditable, and on the record.
Whether that answer satisfies regulators will depend on how issuers use the tools, not on the tools themselves. The upgrade makes the rules inside tokenized assets explicit. The responsibility for using them lawfully remains exactly where it was before the fork: with the issuer.
issuer controls over balances inside the B20 standard and somehow the fork itself got the headlines. thats the actual story here
^ right? everyone covered the fork timing, the TEE prover registry is the part that matters long term
^ agreed, everyone yelling censorship missed that the TEE prover registry quietly sets the verification standard for every issuer that comes next
Exactly. Union and Intersect policies sound boring until you realize issuers can compose transfer rules that freeze a balance almost instantly. Who audits the issuer?
the TEE prover registry partially answers that. verification is on-chain and inspectable now. the issuer reputation half is still manual diligence unfortunately
headline says upgrade, real story is issuers getting explicit power over B20 balances. tokenized assets with a built in off switch
an off switch is only scary if you pretend B20 is crypto native. for tokenized treasuries its table stakes, no issuer signs without it
thats the point though. regulated issuers wont touch rwa without those controls, compliance teams need the kill switch before they sign anything
the seizure operation is the part compliance will love and everyone else should read twice. issuer authority to move balances is a huge deal for anything sold as trustless
seizeWithMemo at least leaves an on-chain reason marker, the old burn-and-block flow didnt even do that. progress, extremely low bar tho
seizeWithMemo leaving a reason marker is real progress, but who validates that the memo is truthful is the next fight. an on-chain note is not an audit
a memo that says compliance action proves nothing about the court order behind it. the chain records authority, legality is still off chain
compliance teams will call the freeze power a feature and honestly they are right. question is whether venues list B20 tokens without a giant issuer-risk disclaimer
kill switch framing is doing a lot of work here. union and intersect are composable transfer rules issuers accept when they issue, not censorship. still watching it closely tho
validity transactions waiting on on-chain conditions is the quiet banger here. conditional settlement on an L2 means escrow logic without a third party holding funds
conditional settlement only beats a lawyer escrow if the conditions are auditable. otherwise trust just moved to whoever writes the predicate
Agreed, conditional settlement without a custodian is the sleeper feature. Cobalt basically ships programmable escrow natively on an L2.
conditional settlement without a custodian is what trade finance has wanted for 30 years. letters of credit in solidity, basically
letters of credit in solidity is generous, the predicate writers just become the new custodians. auditable tho, ill give it that
Validity transactions parked until on-chain conditions clear, plus upgrade scheduling in monitoring mode. Base is quietly shipping the plumbing institutions keep asking for.
sepolia one week before mainnet, nodes forced to 1.4.2, clean activation, no drama. boring well-executed ship, more of this please
Scheduled multipliers changing displayed balances without holder signatures will confuse retail more than any seizure. Wait until balances shift overnight and support tickets explode.
balances shifting overnight without holder signatures is also the best phishing pretext ever shipped. support teams are not ready for that wave
the phishing angle is underrated. your balance changed due to a scheduled multiplier is going to trick thousands who never read the memo
nodes forced to 1.4.2, sepolia a week earlier, clean fork, and still everyone argued about seizure. validity transactions parked until conditions clear is the actual escrow primitive