AI-generated code is still frequently failing security checks — with Claude and GPT-6 worst off

Progress is there, but it's mending a small part of a large wound.

Registered Information Security Specialist Afternoon Exam: "Tracing Code Execution" 11 Questions—A Collection of Problems by Descriptive Answer Pattern

About this articleThe afternoon session of the Registered Information Security Specialist exam includes questions that ...

Veracode Finds First Meaningful Gain in AI Code Security Since 2025 as Average Pass Rate Hits 62%

Veracode today announced a jump from 56% to 62% in overall AI-generated code security — the most significant improvement since 2025.
Microsoft

From help desk impersonation to AD CS abuse: Investigating ESC1 and ESC8 intrusion paths

Teams support impersonation led to remote access, malicious runtime execution, and Certificate Services abuse through ...

[Urgent] Critical Vulnerabilities in PHP, Java, Node.js, and Certificate Authorities | A Beginner's Guide to Types of Cyber Attacks and Zero-Day/N-Day Attacks

IntroductionOver the past few days, there has been a series of reports regarding critical vulnerabilities. These affect "PHP, ...
SourceSecurity.com

Veracode: AI-generated code security improves to 62%

Veracode, the pioneer in application risk management, announces a jump from 56 percent to 62 percent in overall AI-generated ...
Tech Times on MSN

GPT-6 cracks COBOL: 97.9% migration accuracy shifts AI bottleneck from code to enterprise

GPT-6 COBOL migration hits 97.9% behavioral fidelity in OpenAI's latest benchmark, a 16-fold improvement over prior models, but enterprise IT teams should understand the gap between benchmark accuracy ...
remio.ai

Claude Opus 5.5 Code Security Ties GPT-6-Sol at 73%, Leaving Verification Essential

Claude Opus 5.5 code security tied GPT-6-Sol at 73% in Veracode’s fall evaluation, raising the benchmark ceiling while ...