Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i) header
Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i) profile picture

Tier 1 KoL ◈ Research Analyst ◈ Content writer ✍️ ◈ Defi Marketer ◈ DM for Collabs

Malacca, Malaysia
Joined May 2024
Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)Photo by Captain George (💙, 🧡) .edge🦭 (@CaptainDef1i)
DeFi data is becoming a lot more than just BTC/USD or token prices. As more assets move on-chain and new vaults, markets, and tokenized assets emerge, the need for reliable, curated data becomes even more important. The next phase of DeFi needs better data infrastructure. @OasisProtocol oasis-network:native
36
9
54
3,149
Why AI Agents Should Not Automatically Receive Unrestricted Access to Their Owners’ Identity Credentials An AI agent that can act for a person does not need the person’s full identity credentials to do so. Giving it unrestricted access creates risks that selective disclosure is designed to avoid. When an agent is authorised to make purchases, book services or complete forms on someone’s behalf, it needs to prove specific attributes at specific moments. It may need to show that the person is over 18, that the person resides in a certain country, or that a business holds a particular licence. It does not need the underlying passport, the full date of birth, or the ability to answer every possible identity question in the future. Once those credentials are available to the agent without restriction, any compromise of the agent becomes a compromise of the owner’s identity documents. The safer path is the same one already used for people. An independent identity provider verifies the owner once and seals the result to the account. When the agent needs to prove an attribute, it presents a zero-knowledge proof of only that attribute. The service receives a yes. It does not receive the document. The agent does not hold a copy of the document either. The proof is generated from the sealed commitment already linked to the verified owner. Nothing in the proof lets the service or a later party recognise the owner across other interactions. This limit matters as agents begin handling more economic activity. An agent with unrestricted access to its owner’s credentials can be instructed, or compromised, to prove attributes the owner never intended to disclose, or to do so in contexts the owner did not authorise. An agent that can only present selective proofs of pre-verified attributes stays within the permissions the owner set. The owner remains the verified party. The agent inherits the ability to prove what is needed, not the credentials themselves. Verified identity for agents works better when the agent can demonstrate eligibility without ever holding the documents that created it. @Concordium concordium:native
16
50
55
242
Why AI Agents Acting for People May Need Access to Verifiable Attributes An AI agent that acts for a person inherits the same limits as the person. If the person can prove they are over 18 without showing a passport, the agent acting for them should be able to do the same. If the person cannot, the agent cannot either. That requirement is already visible in ordinary commerce. An agent told to reorder weekly groceries and add a case of wine can find the wine and fill the basket. It stops when the retailer needs to know the buyer is old enough. The agent’s options are to fetch the person or to hand over a passport the retailer never wanted to store. The order waits. The same friction appears for any purchase, booking or service that carries an age, residency or authority condition. Concordium’s identity infrastructure has supported this kind of check since mainnet in June 2021. An independent identity provider verifies a person or business once and seals the result to the account. A zero-knowledge proof can later confirm a specific attribute without revealing the underlying document or additional personal data. An agent linked to that account can present the same proof. The retailer receives the yes it needs. It does not receive the passport. The agent does not have to expose its owner’s full identity to complete the transaction. As agents begin making purchases and decisions on behalf of people and businesses, the number of these checks will rise. Age-verification rules are already expanding. Agents that can pay but cannot prove eligibility will simply stop at the same points a person would stop if they refused to upload a document. Giving agents access to verifiable attributes, under the same privacy and accountability rules that apply to the owner, removes that stop without requiring the agent or the service to handle raw identity documents. The infrastructure that made selective disclosure possible for people in 2021 is the same infrastructure that lets an agent prove one attribute on a person’s behalf today. @Concordium concordium:native
14
58
69
230
1/ Privacy is not always about hiding everything. Often it is about deciding who needs to see what. A payroll run does not need to be public. A personal balance does not need to be readable by the whole network. A strategy does not need to be on display while it is still open. @OasisProtocol
12
58
77
269
8/ A practical question before the next tool or process: Who actually needs to see this, and who is seeing it only because the system has no other setting? Most of the time the second group is larger than it needs to be. oasis.net @OasisProtocol
1
8
How Protocol-Level Identity Supports Applications That Require Verification When identity is built into the protocol rather than added by each application, any service that needs a verified person or business can rely on the same check. On Concordium, no account exists without a verified person or business behind it. That rule has been in place since mainnet in June 2021. An independent, regulated identity provider performs the original check once. The result is sealed as a commitment to the account. Concordium does not perform the check and does not hold the document. From that point, every application running on the chain can ask for proof of a specific attribute without building its own identity system. An application that needs to confirm a user is over 18, or that a business holds a particular licence, receives a zero-knowledge proof against the sealed value. The proof confirms only the statement that was asked. The application does not receive the underlying document or additional personal data. Because the commitment is at the protocol level, the same verified account can satisfy checks across different applications without repeating the document verification each time. Accountability remains available through a separate process. Linking a name to an account requires court orders, agreement from two of three independent Privacy Guardians, and the identity provider that performed the original check. No single party, including Concordium, can complete the link alone. Applications can therefore require verification without becoming long-term custodians of identity documents, and without removing the possibility of lawful identification when a court requires it. Protocol-level identity means the verification requirement is enforced when the account is created, not re-implemented by every application that needs it. Applications inherit a verified party, selective disclosure of attributes, and a restricted path to full identification. They do not each have to solve those three problems separately. @Concordium concordium:native
13
63
71
220
Why Trusted Identity Information Is Necessary for Reliable Proofs A zero-knowledge proof is only as reliable as the information it proves against. If the sealed value was never true, the proof will still return yes. The mathematics does not check whether the original document was genuine. That check has to happen before the value is sealed. On @Concordium, the check is performed by independent, regulated identity providers. Three currently do this work: Notabene, Digital Trust Solutions and Global FinReg. Each operates under an established regulatory framework. They examine a passport or equivalent document, complete a liveness check, and seal the result to the account. Concordium does not perform this step. The chain holds only the sealed commitment. The document remains with the provider that checked it. The separation is deliberate. A chain that both issued the credentials and recorded the transactions would hold both halves of the link between a person and their activity. Every claim that nobody can connect the two would then rest on a promise not to look. Keeping the verification with independent providers removes that conflict. Concordium cannot open the commitment. The providers cannot see the account activity. Linking a name to an account still requires court orders, agreement from two of three independent Privacy Guardians, and the provider that performed the original check. This is why the source of the sealed information matters. An anonymous issuer or a self-attested value can produce a valid proof, but the proof has no external basis. A regulated identity provider supplies that basis. The proof can then confirm attributes such as age without revealing the underlying data, while the original verification remains traceable through due process if a court requires it. Reliable proofs depend on trusted identity information at the start. Without that step, a zero-knowledge proof confirms only that a statement matches whatever was sealed, not that the sealed value was ever true. @Concordium concordium:native
14
63
70
273
Paper elections went secret for a practical reason. When nobody can see how you voted, nobody can check that a bought vote was delivered. Public onchain voting brings that problem back. Every vote sits next to the address that cast it, so anyone paying for votes can confirm they got what they paid for. Keeping individual votes private until the close removes that easy check. The counting rules stay public, so the result can still be verified. Governance works better when people can vote without an audience. That is the approach @OasisProtocol is built around. Supported by oasis-network:native. #ROSE #OasisNetwork #Sapphire
11
62
86
419
Verification and Identity Disclosure Are Different Processes Protecting personal information does not require abandoning accountability. The two sit in separate processes. Verification establishes that a real person or business stands behind an account. On Concordium this happens once, with an independent, regulated identity provider that checks a document and a liveness check, then seals the result. Concordium does not perform the check. After verification, no account exists without that sealed commitment. That rule has been in place since mainnet in June 2021. Identity disclosure is a different process that can happen later, and only under specific conditions. Linking a name to an account requires court orders, agreement from two of three independent Privacy Guardians (separate law firms that hold key shares, not the data itself), and the identity provider that performed the original check. No single party holds enough information to make the link alone. Concordium cannot open the commitment. The shop or service that received a proof cannot either. This is why privacy and accountability can coexist. A zero-knowledge proof lets a business confirm you are over 18 without receiving your date of birth or document. At the same time, a lawful process exists to identify the person behind an account if a court requires it. The system is not designed so that nobody can ever be identified. It is designed so that identification cannot happen casually or by any one party acting alone. Anonymity would mean no path to identification exists. Privacy means the path exists but is narrow, requires due process, and involves multiple independent parties. Concordium chose the second. Verification confirms a verified party stands behind the account. Disclosure of who that party is remains a separate, restricted process. @Concordium $CCD
15
86
98
391
Why Online Services Need Better Ways to Verify Eligibility An online service that needs to know you are old enough currently has a limited set of options. Most of them require more information than the service actually needs. The common approach is to ask for a passport, a driver’s licence, or a date of birth. Once that document or data is uploaded, the service becomes responsible for storing it. Stored identity documents are a target. When they are breached, the damage is permanent because a passport number and a date of birth cannot be changed the way a password can. The service gains the answer it needed and also gains a liability it did not. The question being asked is narrow. Is this person over 18? In most cases the service does not need the person’s name, full date of birth, nationality, or a copy of the document. It needs a reliable yes or no. Collecting everything else creates risk for both the user and the service without improving the answer. A better approach separates the check from the document. A person verifies once with an independent identity provider. After that, a zero-knowledge proof can confirm eligibility against the sealed result. The service receives the yes it asked for. It does not receive the underlying document or the full date of birth. The document stays with the provider that checked it. The service never holds a copy it can later lose. This matters more as age-verification rules expand. The UK’s Online Safety Act, the EU’s Digital Services Act, and state-level laws in the US all require platforms to confirm age in certain contexts. Meeting those requirements by collecting and storing passports turns every compliant platform into a custodian of sensitive data. Meeting them with selective disclosure keeps the check and removes the unnecessary copy. Online services need a way to verify eligibility that does not force them to become long-term holders of identity documents. The alternative is not less verification. It is verification that returns only the answer required. @Concordium $CCD
17
56
63
2,600
The private market has always been difficult for ordinary investors to access. @IPOGenie is looking to change that. By combining AI-powered deal discovery with tokenized pre-IPO markets, it aims to bring opportunities around companies like SpaceX, OpenAI and Stripe closer to the on-chain world. Making private markets more accessible is an interesting direction for Web3. 👀 Explore: bit.ly/IPOGenie-captaindef1i
17
80
97
673
Most explanations of zero-knowledge proofs treat verification and disclosure as one action. They are not. Identity verification happens once. An independent, regulated identity provider checks your passport and a liveness check, then seals the result as a commitment to your Concordium account. Concordium does not perform this check. The document remains with the provider. The chain holds only the sealed commitment. From that point, no account exists without a verified person or business behind it. That rule has been in place since mainnet in June 2021. Selective disclosure is the separate step that follows. When a shop or site needs to know whether you are over 18, you generate a zero-knowledge proof against the sealed value. The proof confirms only the statement that was asked. The business receives a yes. It does not receive your date of birth, how far over the threshold you are, or which document was used. Nothing in the proof allows a second business to recognise you as the same person. A zero-knowledge proof cannot make the sealed information true. If an incorrect date of birth was sealed at the start, the proof will still return yes for the rest of your life. The proof is only as trustworthy as the identity provider that performed the original check. That is why the separation between the provider and the chain matters. Concordium never holds the document, and no single party can link a name to an account without court orders, agreement from two of three independent Privacy Guardians, and the provider that checked the document. Verification establishes the sealed value. Selective disclosure lets you prove one attribute of it later without opening the envelope. Treating them as the same step is what creates the myth that a zero-knowledge proof either reveals everything or proves nothing. @Concordium concordium:native
14
57
70
1,385
Captain George (💙, 🧡) .edge🦭 retweeted
There’s a new rider in town 🚴 Privana has been training on Oasis, putting in the miles in closed testing. It’s private DeFi, so your trades and your strategy aren’t on display for everyone. We’ve got something for you on Monday. Keep an eye on our socials 👀 @PrivanaFinance
20
28
2
116
5,985
Town Hall #6 put a clear sequence on what Concordium has been building. Verified humans first. That rule has been in place since mainnet: every account needs a verified person or business behind it, checked by an independent Identity Provider, with identity revealable only under due legal process. Verified payments followed. Verified agents arrived in 2026 with the Agent Registry. The stated next step is trusted agentic commerce in 2027. The numbers shared on the call show the agent layer is already in use. Registered agents reached 5,040. Agent-related transactions rose 47 times to more than 738,000 since July. Six platforms now read the registry. The Verified by Concordium badge works across more than 200 chains. Agents do not have to move. A counterparty can still check that a verified owner stands behind them. The owner’s personal data stays off-chain. The badge means accountable, not endorsed. What was lined up for the months ahead follows the same path. EURQ, a MiCA-licensed euro from Quantoz Payments, is on testnet with mainnet targeted for December. 75 million concordium:native is set aside for builders who make AI agents accountable, invite-led so it follows working products. Protocol-Level Locks go to public testnet in mid-December and mainnet in January. The hour was a statement of where the work sits: agents can already pay. The missing piece for institutions is still who answers for them. Recap: concordium.com/article/conco… @Concordium concordium:native
11
53
67
1,046