Use of Gender-Inclusive Language in News Media

The primary responsibility of the media is to deliver accurate and responsible information to viewers, listeners, and readers. Language plays a crucial role in this process because it does more than communicate information; it also shapes people’s thoughts, social attitudes, and perceptions of the roles of women and men. Therefore, the use of gender-inclusive language in news reporting is an important means of ensuring fairness, inclusion, and equal representation.

Artificial intelligence (AI) is increasingly being used in news writing, editing, translation, and data analysis. It can help identify potentially gender-biased words and suggest alternative expressions. However, AI can also reproduce and reinforce biases present in the data used to train it. Consequently, the use of AI to promote gender-inclusive language in the media must be cautious, context-sensitive, and subject to human oversight.

Media language and gender bias

Language reflects social realities while also shaping how people understand them. Conventional ideas about the social status, professional identities, leadership, and responsibilities of women and men can be reflected in media language. At the same time, such language can reinforce existing inequalities.

Gender stereotypes are simplified beliefs about the characteristics and social roles of women and men. For example, women are often associated with beauty, emotion, cooperation, and caregiving. Men, by contrast, are frequently portrayed in connection with power, strength, competition, and leadership. Repeatedly reproducing these assumptions in news coverage can create a one-sided understanding of gender roles.

One example of linguistic bias is the use of the generic masculine, in which masculine nouns or pronouns are used to refer to everyone, regardless of gender. Research suggests that this practice can encourage readers to form male-centered mental representations.

Differences in the adjectives used to describe women and men are particularly common in reporting on professional roles. Gender bias can also be found in political news. Female politicians are often discussed in terms of their appearance, private lives, and family responsibilities, whereas leadership and authority tend to be associated with masculinity.

Linguistic structure is also important. English, German, French, and Slovak have different rules for expressing gender. Consequently, a strategy that works effectively in one language cannot necessarily be applied directly to another.

Gender-inclusive language and its core principles

The purpose of gender-inclusive language (GIL) is to reduce unnecessary gender bias in language and represent every person accurately and respectfully. When a person's gender is unknown or irrelevant to a news story, journalists should avoid making assumptions about their identity.

Several important principles guide the use of gender-inclusive language in the media.

1. Use gender-neutral terms: Professional titles should avoid referring to a particular gender. For example, Chairperson or Chair can replace Chairman, Police Officer can replace Policeman, and Flight Attendant can replace Stewardess.

2. Avoid the generic masculine: When gender is unknown or irrelevant, journalists can restructure sentences rather than use He or Man as generic terms. Where appropriate, the singular They can also be used.

3. Omit unnecessary gender identification: Expressions such as Female Doctor or Male Nurse should generally be avoided when gender is not relevant to the story.

4. Use equitable forms of address: Forms of address should be applied consistently across genders, while respecting individual preferences. For example, Ms. may be used when a woman's preferred form of address is unknown, as it does not indicate marital status.

5. Avoid stereotypical descriptions: Women should not be unnecessarily portrayed solely as victims, mothers, wives, or caregivers. Similarly, irrelevant emphasis on their appearance or emotions should be avoided.

In public or workplace settings, gender-inclusive alternatives such as People, Everyone, Staff, and Team can be used instead of expressions such as Ladies and Gentlemen. However, language choices must take account of the context of the story, the characteristics of the language, and factual accuracy.

Implementing gender-inclusive language through artificial intelligence

AI can help newsrooms identify potential bias, suggest alternative wording, and analyses patterns of representation in news coverage. Its effectiveness, however, depends on the quality of its training data, its ability to understand linguistic and cultural contexts, and the extent of human oversight.

Research has shown that AI-based language models can associate different professions with conventional gender stereotypes. An analysis of GPT-4o published in 2025 found a tendency to associate healthcare professions more closely with women and physically demanding or technical professions with men. Comparative research involving GPT-4, GPT-4o, PaLM 2, and Gemini 1.0 Pro has also identified evidence of occupational gender bias.

For this reason, simply instructing AI to use gender-neutral words is insufficient. It must also be able to assess when gender is relevant to a news story and when it is not.

1. Real-time editing and language review

AI-assisted editing can identify potentially biased language and suggest alternatives. Media organisations can develop their own gender-inclusive style guides and integrate them into AI-assisted editorial workflows.

Such guides can include preferred terms, discouraged expressions, rules for pronoun use, and examples of the organisation's editorial language conventions. This would enable AI systems to follow not only general linguistic principles but also the specific editorial policies of individual news organisations.

2. Auditing gender bias in news

AI can be used to identify gender bias by analysing news reports before or after publication. This process can examine associations between gender and professional roles, pronoun usage, descriptive language, and patterns of representation across different groups.

However, two types of errors must be considered during such audits. A False Positive occurs when neutral language is incorrectly identified as biased. A False Negative occurs when actual bias goes undetected. Therefore, gender-specific words should not automatically be treated as biased; they must be assessed in context.

3. Increasing diversity in news sources

AI can also help analyse the representation of women and men among experts and the frequency with which their views are quoted in news reports. Newsrooms can develop databases of qualified experts and use AI to identify whether the same individuals are repeatedly selected or whether particular groups are consistently overlooked.

The objective is not to require a fixed number of female or male experts in every news story. Rather, it is to identify imbalances that may arise unintentionally during editorial decision-making.

4. Clear instructions for generative AI

AI systems can be given explicit instructions when generating, editing, translating, or summarising news content. These instructions may include avoiding assumptions about a person's gender, avoiding stereotypes in professional descriptions, using neutral language when gender is unknown or irrelevant, and preserving gender-related information when it is factually necessary.

Such instructions can help reduce bias. However, instructions alone cannot guarantee completely unbiased results. Human editing and review remain essential.

An example of context-sensitive editing in a news headline

British news media “ITV News” on 12 March 2026, published a headline that read: “Woman found dead named as ‘devoted mum-of-two’ as man charged with murder”

The headline describes the deceased woman through her family role, identifying her as a mother of two, and uses the adjective devoted. However, it is important to consider whether these details are necessary to understand the central facts of the murder case.

AI-assisted editing could flag the phrase for review rather than automatically remove it. Editors could examine whether the woman's identity as a mother is directly relevant to the incident, whether the adjective appears in a direct quotation, and whether the information is essential to the news story.

If the family description is not relevant, the headline could be rewritten in a more neutral way, such as: “Woman found dead in Ipswich as man charged with murder.”

This example demonstrates that gender-inclusive editing does not simply mean removing words. It means examining why particular information is included in a news report and whether its use is justified.

Limitations and ethical considerations

Several limitations are particularly important when using AI. First, biases present in training data can be reproduced in AI-generated outputs. Second, AI may misinterpret cultural expressions, pronouns, or the significance of gender in a particular news story. Third, excessive correction may lead to the removal of essential information. Gender can be a crucial factor in reporting on gender-based violence, healthcare, employment, sports, and demographic statistics.

For these reasons, journalists and editors must review AI-generated suggestions before publication. They must preserve the accuracy of the original information, the integrity of quotations, and the context of the events being reported. It is also important to document AI-use policies, regularly evaluate system performance, and maintain human accountability for editorial decisions.

The use of gender-inclusive language in the media is directly connected to representative coverage, social inclusion, and responsible journalism. Language shapes people's perceptions of gender roles, leadership, and professional expertise. News organisations should therefore review unnecessary gender-specific labels, stereotypical descriptions, and male-centred language. AI can support this work through real-time editing, automated bias audits, analysis of diversity among news sources, and the development of appropriate instructions for generative AI.

In conclusion, promoting gender-inclusive language in news media requires more than technological innovation. While artificial intelligence can help identify and reduce gender bias in language, its effectiveness depends on responsible implementation, human oversight, and clear editorial guidelines. Ultimately, the goal is to ensure that journalism represents all genders fairly, accurately, and respectfully.

Journalist and guest faculty in Canadian University of Bangladesh and State University of Bangladesh