PulseAugur
EN
LIVE 01:10:38

Study reveals Anthropic's Opus 5.5 still has thousands of AI writing tics

By PulseAugur Editorial · [1 sources] ·

A study by the marketing agency Graphite, led by Gregory Druck, has identified specific writing patterns, or "tics," that can help distinguish AI-generated text from human writing. The latest analysis focused on Anthropic's Claude Opus 5.5, finding it has significantly reduced its use of em dashes and "affected prose" compared to its predecessor, Opus 5. However, Opus 5.5 still exhibits thousands of distinct tics, such as the over-repetition of phrases like "this matters," which appear far more frequently than in human-written text. The study suggests that relying on AI detectors is unreliable, and a more effective method involves looking for generic phrasing and a lack of concrete detail. AI

IMPACT This research provides methods for identifying AI-generated text, potentially impacting content authenticity and detection tools.

RANK_REASON The cluster details a study analyzing the writing characteristics of an AI model, which falls under research into AI capabilities. [lever_c_demoted from research: ic=1 ai=1.0]

Read on dev.to — LLM tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Study reveals Anthropic's Opus 5.5 still has thousands of AI writing tics

How we ranked this

Signal score
38 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The cluster details a study analyzing the writing characteristics of an AI model, which falls under research into AI capabilities. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. dev.to — LLM tag TIER_1 English(EN) · kalkazu.org ·

    How to recognize AI-generated text: Opus 5.5's writing tics

    <p>You open an email, a post, or a school assignment and something sounds off. The grammar is correct, the sentences are well-arranged, and yet you suspect you're looking at AI-generated text. The hard part is saying why.</p> <p>On October 1, 2026, an agency called Graphite publi…

RELATED ENTITIES

RELATED TOPICS