- OpenAI enables an option to reduce or remove the em dash in ChatGPT when the user indicates it.
- The setting is managed from the custom instructions and is not active by default.
- The measure comes in response to months of criticism over the excessive use of the dash, associated with AI-generated texts.
- Direct impact on media, education and marketing teams in Spain and Europe, with more control over tone and punctuation.
A small but important change is coming to ChatGPT: OpenAI allows that, if the user requests it, The assistant should reduce or avoid using the em dash in their responsesThe company is thus responding to a recurring complaint from the community, where This symbol had become a kind of hallmark of AI-generated texts..
Own Sam altman X acknowledged that it is now possible to tell the model not to use stripes through the personalized instructionsIt's not a default change, but a configurable preference that gives more control over style, especially useful for those who write in Spanish from Spain or work with professional content.
What has changed and how to activate it
OpenAI has adjusted the behavior of ChatGPT to respect an explicit user instruction: If you are asked to avoid the dash (—), the model will moderate or eliminate its presence in the textThe option is integrated into the preferences system, so you don't have to repeat the instruction every time you start a conversation.
To achieve this, simply activate and edit the personalized instructions and add a clear style note (for example, “avoid the em dash and opt for commas or semicolons”). The assistant then adjusts its punctuation and maintains the same criterion throughout different responses, something especially valuable in flows of marketing, education, or technical documentation.
The nuance is relevant: The line doesn't disappear automatically or universallyThose who need it for dialogues or parenthetical remarks can continue to use it; Those who prefer to avoid it will now have a persistent setting to do so.It is an intermediate solution that prioritizes style customization over a blanket ban.
This change comes after months of public debate, memes, and evidence shared in forums showing that, even when requested, the model kept “slipping” the lineWith the new adjustment, the tool aligns better with the preferences of those seeking more natural writing in Peninsular Spanish, where the Overuse of the em dash can result in a forced or repetitive effect..
Why was the em dash overused, and what about AI detection?
The most widespread explanation points to a mixture of training and feedback: Great models learn from books, articles, and forums where the line appears frequentlyFurthermore, some users appreciated its versatility in linking ideas, which would have reinforced its use. Over time, The resource became ubiquitous and, for many, unnatural..
La The proliferation of the “ChatGPT script” turned the dash into an indication of possible artificial authorshipEmails, academic papers, and social media posts came under intense scrutiny when they accumulated em dashes, something that It affected the perception of authenticity even of texts written by people that they simply preferred that sign.
With the new option, One of the most visible clues for detecting AI is losing strengthThis does not mean that artificial content is undetectable, but it does require going beyond such a superficial marker as the insistent use of the stripeIn European academic and publishing environments, this can lead to more robust verification methods. focused on content analysis and not just scoring.
In Spain, where the orthotypography of the dash has clear rules, the ability to control it helps maintain a consistent style with the manuals of media outlets, universities, and government agencies. The practical result: texts closer to the local standard and less "marked" by tics of generative models.
Impact on businesses, media and education in Spain and the EU
For startups, agencies, and editorial teams, having fine control over punctuation improves the brand consistency and reduces the risk of texts “sounding like AI”. It also facilitates the integration of ChatGPT into European workflows with specific linguistic requirements, from internal style guides to regulatory documentation.
In marketing and customer service, eliminating the default em dash prevents writing from looking unnatural on high-visibility channels (website, email, social media). In education, it provides clarity: teachers and students can establish usage criteria and evaluate work with a more focused approach. content and argumentation than in typographic signs.
Quick tips to get the most out of it in your daily life: define in the personalized instructions your punctuation policy (no dashes; prefer commas and parentheses); add a tone note (formal, informative, or journalistic); and internally document examples of acceptable texts to standardize criteria between teams.
A useful detail: if you need the line in any project (for example, in literary dialogues), creates a separate instruction profile. This allows you to switch between styles without reconfiguring each conversation and prevents the assistant from reverting to unwanted patterns. successive departures.
OpenAI's decision does not eliminate a valid and useful sign, but it does offer a precise control Regarding its appearance. For the Spanish-speaking ecosystem in Europe, where stylistic nuances carry weight, the novelty represents a practical improvement: fewer unintentional signs of "typewriter text" and more room for the voice of each organization sound as it should.
I am a technology enthusiast who has turned his "geek" interests into a profession. I have spent more than 10 years of my life using cutting-edge technology and tinkering with all kinds of programs out of pure curiosity. Now I have specialized in computer technology and video games. This is because for more than 5 years I have been writing for various websites on technology and video games, creating articles that seek to give you the information you need in a language that is understandable to everyone.
If you have any questions, my knowledge ranges from everything related to the Windows operating system as well as Android for mobile phones. And my commitment is to you, I am always willing to spend a few minutes and help you resolve any questions you may have in this internet world.