EVERY AI CODE REVIEW VENDOR BENCHMARKS ITSELF, AND WINS

Summary from the TLDR Dev newsletter

AI code review vendors frequently publish self-designed benchmarks where they outperform competitors, but these evaluations often rely on small datasets or synthetic bugs.