Code Metal secures $125M to tackle trust in mission-critical AI
Code Metal, an emerging player in advanced software verification, has raised $125 million in fresh funding to accelerate the use of neuro-symbolic AI in mission-critical software. The investment underscores growing demand for tools that can make complex, AI-infused systems more transparent, auditable and safe.
While conventional machine learning has powered breakthroughs in automation, its opaque decision-making has limited adoption in highly regulated domains such as aerospace, defence, healthcare and financial infrastructure. Code Metal aims to close this trust gap by combining data-driven models with formal, logic-based reasoning.
Neuro-symbolic AI: blending learning with logic
The company’s platform uses neuro-symbolic AI to analyse source code, system specifications and runtime behaviour. Neural components detect patterns and anomalies at scale, while symbolic engines apply formal verification, constraint solving and rule-based reasoning to prove or disprove critical properties.
This hybrid approach is designed to answer questions that traditional AI algorithms struggle with, such as: “Can this control system ever enter an unsafe state?” or “Under what conditions can this transaction workflow deadlock or leak data?” For safety- and security-critical systems, those guarantees can be as important as performance.
Targeting aerospace, defence and regulated industries
Code Metal is positioning its technology for organisations that must demonstrate compliance with strict standards, including avionics software certification, military systems accreditation and medical device regulation. By generating machine-checkable proofs, structured reports and traceable explanations, the platform aims to make audits faster and more reliable.
Investors are betting that as more infrastructure becomes software-defined and AI-enabled, demand will rise for tools that can not only test but mathematically reason about complex codebases. If Code Metal can scale its approach, neuro-symbolic AI could move from research labs into the core toolchains of engineers building the next generation of critical systems.