Can AI generated code be trusted in safety critical systems? 🔐
AdaCore’s GNAT Foundry takes a different approach to AI assisted development by combining code generation with formal proof, mathematically checking whether the resulting software meets its requirements.
It could offer engineers a way to use AI while maintaining the rigorous verification needed in high assurance applications.
Read more on eeNews Europe:
https://t.co/6DRQT2EaTv
The adoption of #formalmethods is a question many engineering leaders are weighing, and today's announcement from @AdaCoreCompany and @cimarroninc offers a practical example: https://t.co/weLN0eW9Qf
@AdaCoreCompany is collaborating with Carnegie Mellon University researchers behind Verus, an open-source tool for mathematically proving the correctness of Rust code. https://t.co/QtewUwCmmF
Are you finding joy with AI development? Check out the new @AdaCoreCompany blog to discover how one of our engineers has been experimenting by using AI models with Ada and SPARK. https://t.co/c4WlYWZw1A
Formal software verification with AI is getting attention: Anthropic recently published research on using LLMs for formal verification. View @AdaCoreCompanys latest experiment...https://t.co/XtWqiIq6l3
We need to get better and better at defending our codebases. CodeSonar for Ada (formerly GNAT SAS) can be a huge help for securing software as its SAST (Static Application Security Testing) features, such as Taint Analysis https://t.co/B2xY83Blbv
@AdaCoreCompany
At #NVIDIAGTC Berlin, European mobility leaders showcased how NVIDIA Hyperion supports safe autonomous driving and robotaxi fleets. We also introduced Halos SAFE to accelerate AV safety validation, and Alpamayo 2 Edge for in-vehicle reasoning.
Read more: https://t.co/Zs9bBAK7Jq
Sponsored - Embedded Executive LIVE at embedded world North America: AdaCore
Rich Nass and Mark Hermeling of @AdaCoreCompany talk tech at @embedded_USA 2026. #ewna#AI
https://t.co/8wq33IQxPo
Preparing for the #CRA? Start with your engineers. The CRA asks whether your team applies secure development consistently, not whether a policy exists. Read the new blog from @AdaCoreCompany https://t.co/K2iFEvIUuM
The main themes of the day so far at the @HISC_conf are speed and acceleration. How can AI be safely used in a high integrity systems? Recently @AdaCoreCompany released the GNAT Foundry which answers this! https://t.co/F2LKONmO2C
@AdaCoreCompany is attending the High Integrity Software Conference (HISC) 2026 in Birmingham on October 13, where it will exhibit and present a session on agentic AI workflows and software verification for high-integrity systems.
https://t.co/Bzn5Xf5f2b
A new venue for @HISC_conf that takes place tomorrow, it's looking really fantastic. The Vox Birmingham is such a super-bright space with good acoustics, too (which, as a partially deaf person, I really appreciate). Thank you @ZiaBiaEvents :)
In September, @AdaCoreCompany launched GNAT Foundry which demonstrates our approach, not to treat the AI output itself as trustworthy, but to surround it with deterministic tools that generate independently checkable evidence. Free webinar next week https://t.co/WQfrLLwAr3