BREAKING
OpenAI: 30% of SWE-Bench Pro Broken
0
tasks
0
repositories
0%
broken tasks
How Tasks Break Down
1Hidden requirements
2Contradictory instructions
3Overly strict tests
4Incomplete grading
Two Benchmarks, Same Fate
SWE-bench VerifiedDropped Feb 2026
500 human-verified tasks
Dropped over contamination
SWE-Bench ProDropped Jul 2026
1,865 tasks by Scale AI
Dropped: 30% broken
The Benchmark Cycle Repeats
Reliable Evaluation Stays Hard
AI NEWS BLITZ
OpenAI says nearly a third of SWE-Bench Pro tasks are broken and drops its endorsement.