This site is fictional demo content. It is not real news or affiliated with any real organization. Do not treat it as fact or professional advice.

NextPaper Logo
nextpaper.ccCYBER CHRONICLE
RSS
EN中文

Full article

FULL TEXT

View this issue
BriefAI

AI Program Verifier DeepCodeForm Launches: Automatically Proving Software Code Is Bug-Free

Microsoft Research has released DeepCodeForm, a formal verification AI system that can automatically prove the functional correctness of large software systems within hours, having successfully verified 120,000 lines of critical Windows kernel code as bug-free

AI Program Verifier DeepCodeForm Launches: Automatically Proving Software Code Is Bug-Free

Microsoft Research today released the DeepCodeForm formal verification AI system. The tool combines large language models with automated theorem provers to automatically prove the functional correctness of large software systems within hours.

Traditional formal verification requires human experts to manually write verification specifications and proofs; a medium-scale module could take months. DeepCodeForm uses AI to automatically infer expected behavioral specifications from code and documentation, then invokes SMT solvers like Z3 to generate mathematical proofs.

Kai Zhang, Principal Researcher at Microsoft Research, said: "DeepCodeForm has successfully verified the memory safety and concurrency correctness of 120,000 lines of safety-critical code in the Windows kernel. The entire process took 14 hours — if done by human experts, it would have been estimated at 18 months."

The system is already available as a beta feature in GitHub Copilot Enterprise, where developers can automatically trigger formal verification during the code review phase.