The new tools aim to make AI’s “Vibe Coding” safer for cryptocurrencies



short

  • The ASI Alliance and Matterhorn said they are building tools to reduce risks from AI-generated blockchain code.
  • The platform combines automated analysis, human review, and testing tools to audit smart contracts before deployment.
  • The partnership runs on ASI:Chain and aims to have 20,000 developers on board in 2026.

Artificial intelligence has begun writing the code that moves money on the blockchain. The Super AI Alliance and development platform Matterhorn say they want to make sure this code is secure.

The Matterhorn and the ASI Alliance on Friday announced the new initiative focused on “Atmosphere coding“- A feature of the Matterhorn platform that allows developers to describe the application in plain language, and the AI ​​instantly generates the complete smart contract code. While this technology speeds up development and reduces barriers to building applications, it also introduces the risk of AI creating flawed or insecure code that attackers can exploit.

“We are at the beginning of a world where decentralized applications become just apps, as common as the websites and apps we use today,” the company said in a statement. “Every other tool in this space is racing to ship code faster. We think this is the wrong race. Creators who build decentralized applications that handle real money and real users need a platform they can trust, and this partnership is how we’re building it.”

To mitigate this threat, Matterhorn founder Abhinav Ramesh said the company is working with third-party security auditors and automated tools to help developers review AI-generated smart contracts before deployment.

“We partner with security auditing companies who can provide auditing services through Matterhorn to builders on Matterhorn,” Ramesh said. Decryption. “We also have AI agents that do proxy audits, but we would never recommend doing this just for mainnet applications.”

The Matterhorn development platform is designed to integrate with it Asi: seriesa blockchain network developed by the AI ​​Super Alliance, is a decentralized artificial intelligence Collective It includes Fetch.ai, SingularityNET, and CUDOS, giving developers a single environment to build, review, and deploy decentralized applications.

“We make it easier for users to connect MCPs, build/use skills, build dApps, and deploy from a single platform,” Ramesh said. “We are working with the ASI team on ‘blessed models’ to facilitate the creation of more secure contracts specifically for formal verification-based languages.”

The company said developers can communicate with third-party validators through the platform before launching contracts on the live blockchain. However, while Matterhorn said its platform speeds up the development process, it does not guarantee security.

“We are a powerful enabler for creators who want to build on Web3,” Ramesh said. “There are absolutely no guarantees of any kind from the Matterhorn team regarding safety and security.”

Matterhorn and the ASI Alliance are developing “blessing models” to help developers build more secure smart contracts while integrating ASI:Cloud to provide computing power to AI systems that generate and analyze code for MeTTa, the ASI:Chain programming language, Ramesh said.

Partnership comes as Artificial intelligence agents They are increasingly moving into the cryptocurrency industry, as developers experiment with systems that can manage them governorAnd execute trades and carry out financial tasks on-chain, which has led to the emergence of new tools and research aimed at controlling Risks When those independent systems handle Cryptocurrency.

Khylar Crawford, chief innovation officer at SingularityNET, said much of the blockchain industry relies on a “patch and pray” approach — writing smart contracts in languages ​​unsuitable for complex concurrency and relying on validators to detect flaws — while F1R3FLY and ASI:Chain use what he called a “patch-by-construct” architecture based on Rho calculus.

“We’re not guessing whether an app is safe or not, we’re proving it mathematically using spatial behavioral types,” Crawford said. Decryption. “Before a single line of code touches the live network, the same mathematics ensures that there are no deadlocks, exploit race conditions, or money leaks.”

Daily debriefing Newsletter

Start each day with the latest news, plus original features, podcasts, videos and more.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *