Study identifies six linguistic patterns that make AI text hard to read
What happened
On 27 October 2026, researchers at New York University and the University of Texas at Austin published a study summarising six linguistic patterns that make AI text hard to read. The study had 14 models, including Claude, GPT-5.6, GLM-5, Kimi K2.5, Llama and Olmo, generate everyday conversations, coding sessions and argumentative essays, analysing 120 conversations, 45 coding sessions and 980 essays and comparing them with human texts such as film subtitles, StackExchange posts and student essays. Reports so far disclose only the study design and sample size, not the specific content of the six patterns.
Written by AI from the reports · updated 1 h ago
Timeline
Follow the reports to see the event from each side.
- The DecoderResearchers find six patterns that make AI text so exhausting to read
A study from New York University and the University of Texas at Austin shows that frontier language models systematically overuse linguistic patterns that are hard for humans to parse. The researchers trace the problem to training on code and math, where correctness gets rewarded but clarity doesn't. Smaller models, by contrast, write more like humans.
Heat of this event
Not enough continuous observations yet to draw a trend.