Most AI news products ask readers to trust the feed. Trust that a summary is accurate. Trust that it has not been rewritten. Trust that restricted content is handled responsibly.
NewsAgents takes a different route.
Built by an independent developer on the Concordium protocol, NewsAgents shows what happens when AI content, user eligibility, and agent ownership are backed by proof, not empty platform promises.
A reader connects a Concordium wallet, chooses topics, and assigns AI Agents to fetch and summarise news. Every summary is anchored on-chain.
A Feed You Can Check
An AI summary can change after publication. A word changes. A claim softens. A sentence disappears. The reader sees the new version with no record of the old one.
NewsAgents replaces that promise with a check.
Each summary is hashed when published. That hash is written to the Concordium blockchain. Before a story opens, the site checks the current version against the on-chain anchor.
This does not prove that every summary is correct. It proves something narrower: the version a reader sees is the same version that was anchored when published.
If a single word has changed, the hashes no longer match. The story does not open. That is the difference between a feed that asks for trust and a feed that can prove its integrity.
“We didn’t want readers to take our word for it. Building on Concordium means they don’t have to. The chain holds the proof, not us. Every summary our AI Agents publish can be checked against its anchor, and the reader’s identity never touches our servers.” said a developer from the team behind the NewsAgents.
Real Identity, Zero Data Exposure
Some topics carry an age restriction: gambling, alcohol, and the adult industry. The usual answer is to collect more personal data: a date of birth, document scan, or photo ID.
NewsAgents does not need to collect that data.
The reader signs in with a Concordium wallet, so every account is backed by a verified human. To access age-restricted topics, the readers age is checked through a zero-knowledge proof without revealing any additional information.
The platform does not need the reader’s name, document number, or date of birth. Concordium does not put that data on-chain either. There is no new pool of user identity data for the application to hold, monetise, or lose.
This is privacy with accountability, not anonymity.
AI Agents With an Accountable Owner
The reader’s AI Agents are registered on the Agent Registry and signed once by the reader. They are agents with an on-chain identity, assigned by a verified person, working on that person’s behalf.
The user owns the agents. The agents carry identity. The work they produce can be checked. The human behind them remains accountable.
That is why the protocol matters. A private database can be changed inside the platform. A public anchor gives the reader something external to check against.
NewsAgents shows what Concordium’s identity layer makes possible outside payments: private eligibility, checkable content, and AI Agents that can act for verified people without becoming unaccountable black boxes.
If a news desk can prove its integrity to a reader, what else can an AI Agent prove once it carries a verified identity of its own?
Test for yourself: https://newsagents.online/