ConVer: Using Contracts and Loop Invariant Synthesis for Scalable Formal Software Verification
arXiv:2605.27051v1 Announce Type: cross Abstract: Formal verification of large C programs is impeded by state-space explosion: Bounded Model Checking (BMC) tools must encode the entire state space up to the predetermined bound by unrolling all nested constructs. We present ConVer, a top-down compositional verification tool. Given a C program with a top-level assertion, ConVer decomposes verification top-down: it uses a large language model (LLM) to synthesise function contracts from the system property, then alternates system-level and function-level checks in a CEGAR-CEGIS loop, refining cont
The paper leverages recent advancements in large language models to address a long-standing challenge in formal verification, integrating AI to overcome previous scalability limitations.
This development suggests a pathway to more reliable and secure software, particularly for complex systems, by reducing the burden of manual verification and enabling its application to larger codebases.
The ability to formally verify large C programs using AI-assisted contract and loop invariant synthesis significantly lowers the barrier for comprehensive software assurance and could accelerate the adoption of formal methods.
- · Software developers
- · Cybersecurity industry
- · Critical infrastructure sectors
- · AI software tool developers
- · Malware developers (long term)
- · Manual software testers
- · Companies with poor software QA
Increased adoption of formal verification techniques in software development due to improved scalability and automation.
A potential reduction in critical software bugs and vulnerabilities across various industries, leading to enhanced system reliability and security.
New regulatory standards requiring AI-assisted formal verification for high-assurance software, creating a market for specialized tools and services.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI