&Truth launches with $70 million for software verification
The company is developing independently checkable proofs against formal software requirements; its universal verification system remains a development ambition.
More generated code creates a larger verification problem. &Truth emerged from stealth with $70 million in seed capital to help developers assess whether software meets precise requirements.
What happened
The company announced its launch on 29 October. Its approach combines AI with formal methods to produce mathematical proofs that can be checked independently.
It is developing tools for turning human requirements into formal specifications and verifying software across programming languages. The company builds on open-source technologies including Lean and plans to release foundational libraries. A universal verification machine is its development ambition, rather than an established guarantee for every program.
Why it matters
Testing examines selected cases; formal verification can establish properties under explicit assumptions. That could be valuable for critical systems and for reviewing code produced at greater speed.
The specification remains essential: proving the wrong requirement does not make a system suitable for its intended use.
The bigger picture
The practical test is useful coverage at a manageable cost. Developers need understandable specifications, reliable proofs and workflows that fit real codebases. Funding provides resources to pursue that system, while its broad claims still need demonstrated results.