Join our daily and weekly newsletters for the latest updates and the exclusive content on AI coverage. Learn more
We have heard a lot about AI agents – tools fueled by generative AI models that can perform actions without too much supervision or human intervention.
But they are largely a new curiosity for most people, and as far as we can say, very few people trust the AI agents to buy or conclude contracts on their behalf – for the moment.
GenlayA startup just out of stealth, believes having a technology that will provide the component of missing “confidence” to the economy of AI agents.
The idea of Genlayer is an infrastructure powered by the blockchain which allows the agents of the AI to write contracts, to settle payments and to perform agreements independently.
Last fall, the company announced that it had raised $ 7.5 million Notable investors, including Arthur Hayes (Maelstrom), Arrington Capital and North Island Ventures, to give life to this vision.
How to make AI agents trustworthy for people – and to each other
AI agents already prove their ability to analyze data, conclude agreements and manage assets, but there is a fundamental problem: they are not intrinsically trusting. Unlike humans, AI agents do not fear prosecution or reputation damage – so how do they apply agreements?
Albert Castellana, CEO of Yeagerai (Building Genlayer company), considers this as a critical defect in the development of AI today.
“Even in a situation where you have agents who can trade between them, how do they trust?” Castellana said in a recent interview with Venturebeat. “The AI does not sleep, the AI works on a global scale, the AI cannot go to prison. The legal system will have a big problem concerning this type of situation. »»
Traditional intelligent contracts, which blockchain -based transactions are too rigid for AI trade.
They cannot process unstructured data, understand a complex language or adapt to changes in the real world.
Genlayer wants to upgrade intelligent contracts into “intelligent contracts” – more flexible and fed by AI agreements that operate as human contracts.
Unlike traditional blockchains that require external oracles to access out chain data, Genlayer directly integrates AI in the protocol. Intelligent contracts can natively transport live web data, process inputs in natural language and reason on complex and real conditions – all without counting on third -party services.
“Blockchains allow self-application contracts, but they have limits,” said Castellana. “They cannot connect to the outside world, they cannot include unstructured data. But AI needs contracts that look very much like human contracts – quickly, inexpensive and adaptive. »»
Genlayer solves this with “optimistic democracy, a model of consensus focused on AI where several validators – each using different models of languages (LLM) – vote on the fact that a contract or a decision generated by AI is valid. This ensures that no single AI model has control and prevents manipulation.
“We have created a blockchain where validators, even if they obtain different answers from AI or Internet, can still reach a consensus,” said Castellana. “It is essentially a judicial system for the future of trade.”
How does Genlayer approach work
Basically, Genlayer operates as a layer of a -native confidence – an independent system which guarantees that the agents of the AI operate fairly in financial transactions, the execution of contracts and the settlement of disputes.
Key characteristics include:
- Smart contracts: Agreements fed by AI which process natural language and access live web data.
- Decision -making led by AI: A consensus model where several AI models vote on the results to ensure reliability.
- Optimistic democracy: A model of governance based on blockchain which prevents the manipulation of AI using decentralized decision -making.
- Chain interoperability: The possibility of connecting smart contracts with real data and Internet sources.
- ZKSYNC integration: Scalability, low costs and security at the level of ethereum.
At the heart of Genlayer is an “optimistic democracy”, an improved delegated proof model (DPO) which directly integrates AI into the blockchain validation. Instead of relying on deterministic logic, validators connect to LLM to process natural language, interpret data and run complex decisions on the chain.
When a transaction is subject:
- A leader validator deals with demand and offers a result.
- A set of validators recomprate the transaction independently, validating the leader’s proposal.
- If the majority agree, the transaction is finalized. Otherwise, a new leader is selected and the process is repeated.
This mechanism prevents manipulation and guarantees that the decisions generated by AI are supported by consensus rather than by the judgment of a single entity.
Inspired by Condorcet jury theorem – A theory of mathematical and political sciences of the 18th century by the Marquis de Condorcet which says that a jury is more likely to make a correct decision with more participants – the system brings together the results of the AI through several validators, guaranteeing equity and even reliability for non -deterministic tasks such as the interpretation of legal contracts, verification of data on the supply chain Dynamic price.
The approach is described in a white paper Posted by the three co -founders of Genlayer – Castellana, José María Lago and Edgars Nemše. You can find it integrated below.
Why Genlayer thinks that his moment has happened
The breed to create autonomous AI companies is accelerating. Companies like Openai deploy AI agents who can operate independently, but they still count on slow legal and financial systems focused on humans.
“The AI will not wait for lawyers,” said Castellana. “If we want AI to participate in the economy, we need infrastructure that corresponds to its speed.”
Other startups attack Skyfire And Pin – But Genlayer adopts a different approach. Instead of focusing on the construction of AI agents themselves, Genlayer creates the confidence layer that allows them to transform.
“There are 100 startups working on AI agents,” said Castellana. “But confidence requires a third party. We build this third party – the infrastructure that makes the IA trade possible. »»
To encourage validators and cover the costs of the execution of intelligent contracts, Genlayer introduces a native gas token called Gen. Users pay transaction costs in Gen, which are then distributed to validators as a reward for their services.
This approach ensures that AI -oriented transactions remain fast, low -cost and autonomous. In addition, the Genlayer chip -based look model aligns incentives by rewarding honest validators and penalizing bad players with reduction mechanisms.
What is the next step?
Genlayer’s Testnet launches the first projects focused on:
- IA supply chains which negotiate logistics independently.
- AI propulsion influencer marketing where payments are based on performance.
- Decentralized financial applications which use AI for the execution of trade.
- Autonomous autonomous autonomous organizations (DAO) which adjust governance according to the comments of the community in real time.
- Insurance and loan fed by AI which adapt terms based on external risk factors.
The company will present its technology to Conference of the Ethereum community (ETHCC8) Taking place from June 30 to July 3 in Cannes, France.
IA’s economy arrives – but without confidence, it will not extend. Genlayer bet that the contracts directed on the AI, applied to the blockchain, will be the basis of a machine focused on the dollar billion.
“We are building the Legal IA trade legal framework,” said Castellana. “It is a question of allowing the AI to work together, not only faster, but with confidence.”
