Aptos says it is the first L1 to support AI‑assisted formal verification for dynamically scheduled Move contracts, pitching Move Prover as an “oracle” between markets and machine agents.
- Aptos says it is the first Layer‑1 blockchain to support formal verification for dynamically scheduled smart contracts, using AI‑generated specifications and its Move Prover tool.
- The team frames Move Prover as an on‑chain “oracle” that mathematically proves contract behavior, positioning it as a trust layer between markets and machine‑driven trading systems.
- The upgrade lands as Aptos steps up a broader $50 million push into AI‑driven markets and on‑chain finance infrastructure.
Aptos has declared that its network is now “the first L1 to support dynamically scheduled formal verification,” arguing that its Move smart contract stack can be mathematically proven to behave as designed even when contracts use higher‑order functions and dynamic dispatch.
In a post on X, Aptos Labs said that “AI writes the specifications, mathematics proves their correctness, and the Move Prover serves as the oracle — this is the trust layer between the market and the machine,” outlining a workflow in which large language models generate formal specs that are then fed into the Move Prover for automated verification.
Aptos leans on Move Prover to court AI‑driven markets
Expanding on the X announcement, the Aptos team emphasized that “Move is the only smart contract language with native formal verification built in, with extended support for dynamic dispatch now enabled,” a reference to recent work bringing first‑class functions and dynamic dispatch into the Move toolchain while keeping them inside the prover’s threat model.
According to Aptos’ Move Prover documentation, the verifier is already used at protocol level “to verify core logic — staking, metering, code deployment, and supporting data structures,” with a design goal of letting domain experts with a mathematical background but “not necessarily a software engineering background” audit what on‑chain programs actually do.
The latest research on the system, published in a May 2026 paper titled “Formal Verification of Imperative First‑Class Functions in Move,” notes that dynamic dispatch dramatically increases the complexity of the state space and required Aptos engineers to redesign how the prover generates and checks verification conditions for higher‑order Move code.
This push into AI‑assisted verification arrives alongside a $50 million Aptos funding program for on‑chain markets and AI systems, where the company has promised encrypted mempools and confidential perps in a bid to build “institution‑friendly” trading rails that can withstand adversarial bots and the next wave of machine‑driven strategies.
In that context, framing Move Prover as an “oracle” between markets and machines is less marketing flourish than an attempt to sell formal methods as the only scalable way to police smart contracts once AI agents are authoring, deploying, and trading against them in real time.
