OpenAI improves ChatGPT coding ability in low-resource languages like COBOL
OpenAI announced improvements to ChatGPT's ability to code in low-resource programming languages, particularly COBOL and Fortran.
Some disagreement beneath a largely shared set of facts.
Advanced
See who covered this in depth — and who just repackaged it, plus the angle everyone missed.
The free cross-source synthesis is just below.
Argon Synthesis
IncludedOpenAI announced improvements to ChatGPT's ability to code in low-resource programming languages, particularly COBOL and Fortran. According to ITmedia NEWS, the latest model (GPT-6 Astra) achieved an average 87.4% pass rate on engine-building evaluations across 20 languages and reached up to 97.9% accuracy on COBOL-to-modern-language migration tests. NV.ua noted the practical implications: the capability could help modernize legacy software used by banks, government agencies, and critical infrastructure.
Where they diverge
All sources agree on the core achievement. ITmedia NEWS provides the most specific performance metrics (87.4% average, 97.9% on COBOL migration). NV.ua emphasizes the real-world application to legacy system modernization in financial and government sectors, while OpenAI's own announcement frames it as a general advancement in software engineering evaluation.
This synthesis is AI-generated by Argon from the listed sources. It may summarise inaccurately or miss nuance — rely on the original articles for specifics. Read more about Argon's crawler policy.
1 subject in this story
Open any player to see how outlets across the spectrum tend to cover it.
Aggregate substance
2 sources- SensSensationalism
- 4.5
- SpecSpecificity
- 5.5
- CorrCorroboration
- 2.0
- NovNovelty
- 4.5
- IndIndependence
- 0.5
Sources spread by 25 pts
Coverage map
2 sourcesYou're seeing this mostly through a Neutral · OpenAI lens.
No coverage from: UK · EU · Rest
How the sources framed it
The same story, grouped by the stance each outlet took.