ETHICS AND GUIDELINES

As a part of ION DIGITAL LTD, ionnews.mu and ION NEWS digital platforms adhere to the following Editorial Ethics & Guidelines:

ION NEWS has evolved and will continue to evolve as it builds a portfolio of modern editorial networks and partners across multiple mediums driving the future of journalism and entertainment. However, our commitment to the core values of integrity and passion will never change.

We believe in working with talented people, the judgment of our staff, and the transcendent importance of serving the interests of our audiences. Through our editorial guidelines we aim to give our teams clear guidance about what to avoid and the public knowledge of what to expect.

Simultaneously, we recognise the impossibility of reducing the complexity of real life to a simple checklist or rulebook and encourage conversation and dialogue with colleagues and supervisors about concrete situations as superior to trying to craft an ethics policy that would address every conceivable dilemma.

As the needs of our newsrooms and audiences change, our guidelines will adapt in kind to ensure that ION’s work is always deserving of our audiences’ trust.

Advertising and Revenue Partnerships

ION NEWS is home to editorial networks and the businesses that enable them. In no case, however, do these efforts interfere with the integrity of our editorial teams or the content they create.

This includes ION NEWS and ION LABS’ divisions responsible for selling advertising and editorial sponsorships on our properties; advertising across the publisher-led marketplace.

Advertisements do not necessarily reflect the views of ION DIGITAL or our editorial teams.

Conflicts of Interest, Activism and Disclosures

ION NEWS’ newsroom has complete editorial independence and our editorial staff and contributors take their beats seriously. Any editorial team member or contributors will be recused from a story and/or publicly disclose conflicts of interest when editorially appropriate. This may include a personal or family relationship, personal financial investment, or relevant political activity. Our staff will also refrain from demonstrating or donating to causes that could compromise their journalistic coverage of an issue.

Corrections, Updates and Deletions

ION NEWS makes every effort for content to be completely accurate upon publication. If a correction is required, however, we are transparent and update a story if new information either adds to or invalidates the original story. We will issue corrections for a factual error or if a typo in the copy could cause audiences to misunderstand the story.

We almost always leave all editorial content live, with notes and corrections as needed. We may, in rare instances, remove content from our websites, social media pages, or related platforms for legal reasons or extenuating circumstances. Removal will never be at the direction of, or be influenced by, our advertisers and will always be approved by editorial leadership.

Some ION NEWS editorial networks accept financial contributions from individual audience members and funding from organisations. However, this financial support does not affect editorial independence.

Gifts and Giveaways

We do not accept gifts or other consideration from companies as a condition or incentive to write a review or story, whether favorable or unfavourable. Any gifts accepted will be of minor value and in no way compromises or guarantees editorial coverage of the event, product, or service.

Occasionally, ION NEWS will conduct contests and giveaways for our audiences. We will post rules for each contest and giveaway that will be binding for those who decide to participate.

While giveaways or contests may be sponsored by an advertiser or partner, they are not, and should never be, considered endorsements of the entities involved or their products.

Social Media

ION DIGITAL hires our editorial employees for their strong, expert voices, and values their ability to build positive communities and influence on our platforms and their own. We implore all staff to wield their tool/voice with extra care. Behaviour on social media is a reflection of both one’s personal and professional self and, as a voice of ION DIGITAL and our editorial networks, we expect social posts to apply the same journalistic integrity required in our editorial content.

Note: our editorial guidelines leave no room for indulging harassment on social media. All ION DIGITAL employees are encouraged to access and review our protocol for reporting online abuse.

Sourcing

ION NEWS obtains news and content from a variety of sources. Some of our sources do not wish to be named, and we grant anonymity when we know our sources to be credible and reliable, and that their story would otherwise not be told. We make best efforts to confirm the veracity of the information provided by anonymous sources and will not include any information we have not thoroughly vetted. It is our policy not to pay sources.

ION DIGITAL Values

Our values remain our cornerstones at all times. In an industry that is constantly evolving, we will thrive on change and innovate to improve. We will always hold ourselves and our teammates to high standards and build an environment that cultivates passion, benefits collaboration, and respects all voices and identities.

Be inclusive

We believe it’s a moral and business imperative to cultivate diversity. Seek new perspectives. Be inclusive in your work and your behaviour.

Be ambitious

We are ambitious, unapologetically. Embrace your competitive spirit and desire to excel when it counts. Don’t wait for opportunities to come to you. Be resourceful. Make things happen.

Collaborate well

We are inspired by and trust in the people with whom we work. We grow stronger through shared knowledge, resources, and common goals. Work together to make ION DIGITAL better.

Demand quality

We hold ourselves accountable and hold our teammates to high standards. It takes effort and clear priorities to make quality happen. Ask for help when you need it. Provide help when asked.

Take risks

We move quickly to innovate. We take informed risks, learn from our experiments, and talk honestly about our actions. Iterate to improve.

Thrive on change

We stay agile amidst the changing media and tech landscape. Be flexible, fail fast, and keep improving. Don’t fear change — embrace it.

AI POLICY

In brief — what our readers can expect

We use artificial intelligence the way we use any tool in the newsroom: to do our work better, never to replace the judgement of the journalists responsible for it. Three commitments sit above everything else in this document, and we will not compromise on them.

A journalist is accountable for every story we publish. AI never has the final word, and it never publishes on its own.

Nothing reaches our audience unverified. Anything an AI tool helps produce is checked by a person against real sources before it appears under our name.

We are honest about it. Where AI has played a meaningful role in what you read, see, or hear, we tell you.

If you ever have a concern about how AI was used in our reporting, write to us at [email protected].

  1. Why this policy exists

Artificial intelligence is now part of how news is gathered, produced, and distributed worldwide. Used carelessly, it can spread errors, fabricate sources, mislead audiences with synthetic images, and erode the trust a newsroom spends years building. Used carefully, it can help a small newsroom transcribe interviews faster, work confidently across Kreol, French, and English, sift large documents, and free reporters to spend more time on original reporting.

This policy sets out, plainly and publicly, where ION News draws those lines. It exists so that our journalists know exactly what is allowed, and so that our readers, our sources, and anyone who scrutinises our work can hold us to a clear standard rather than guess at one.

  1. Scope

This policy applies to everyone producing journalism for ION News — employed staff, freelancers, columnists, and any contributor — across every platform we publish on, in every language we work in, and to text, images, audio, video, and data alike. It covers any use of an AI system, whether a general tool, a specialised newsroom tool, or one built into software we already use.

  1. Our core commitments

These principles govern every decision in this policy. Where a specific rule below is silent or unclear, staff should return to these.

Human accountability. A named human editor or reporter is responsible for everything we publish. AI is never the author of record and never the final decision-maker. Responsibility cannot be delegated to a machine.

Accuracy above speed. AI tools can produce confident, fluent text that is wrong. We treat everything they generate as unverified until a journalist has confirmed it against trustworthy, original sources. Being first never outranks being right.

Transparency with our audience. We do not hide our use of AI, and we do not pretend AI work is human work or human work is something it isn’t. Where AI has shaped what the audience receives in a way that matters, we disclose it.

Protection of sources and privacy. The trust of a source is sacred. We never expose confidential information, unpublished material, or personal data to AI tools that could store, leak, or learn from it.

Fairness and the avoidance of bias. AI systems learn from data that carries human bias. We stay alert to that, particularly in how people, communities, and languages are represented, and we do not let a tool launder prejudice into our reporting.

Editorial independence. No AI vendor, platform, or automated system sets our editorial agenda. Our judgement about what is newsworthy and how to report it remains entirely our own.

  1. How we may use AI

Used as an assistant and always under human control, AI may support our work in ways such as:

  • Transcription of interviews, press conferences, and recordings, with the transcript checked against the audio before any quote is used.
  • Translation and language support across Kreol, French, and English — to draft a first translation, suggest phrasing, or make our reporting accessible in more languages — with every published translation reviewed by someone who knows the language.
  • Research assistance, such as summarising long documents, reports, or background material to help a journalist find what matters, on the firm understanding that the journalist then goes to the primary source.
  • Copy-editing and proofreading — grammar, clarity, and consistency — without changing the meaning or facts a reporter intended.
  • Headline, summary, and SEO suggestions for an editor to accept, reject, or rewrite.
  • Data work, such as cleaning, sorting, or spotting patterns in large datasets, with findings independently verified.
  • Idea generation — angles, questions, lines of inquiry — as a starting point for human reporting, never as a substitute for it.
  • Accessibility, such as drafting alt-text for images or captions, reviewed before use.

In all of these, the AI assists; the journalist decides.

  1. How we will not use AI

The following are not permitted at ION News:

  • Publishing AI-generated text without human review and editing. No article, or section of one, goes out as the unchecked output of an AI tool.
  • Fabricating anything. We never use AI to invent quotes, sources, people, events, statistics, or details, and we never present AI-generated material as something a real person said or did.
  • Passing off synthetic images, audio, or video as real journalism. AI-generated or AI-altered visuals are never presented as genuine photographs, footage, or recordings of real events.
  • Creating deceptive depictions of real people — “deepfakes” or manipulated media that could mislead the audience about what a real, identifiable person said or did.
  • Feeding confidential or sensitive material into external AI tools — including unpublished stories, the identity or information of confidential sources, leaked documents, or personal data — where the tool could retain or expose it.
  • Fully automated reporting on sensitive matters, such as deaths, crime, courts, elections, communal or religious matters, or anything affecting vulnerable people, without direct human reporting and oversight.
  • Using AI to surveil, profile, or unmask individuals in ways that would be unethical for a journalist to do directly.
  1. Human oversight and sign-off

Every piece of journalism that involved AI in a meaningful way passes through a responsible human editor before publication. That person is accountable for its accuracy, fairness, and compliance with this policy, exactly as they would be for any other story. If a tool’s contribution cannot be verified, it is not used.

  1. Verification and accuracy

AI output is a lead, not a fact. Names, dates, figures, quotes, and claims that originate from or pass through an AI tool are confirmed against original, trustworthy sources before publication. Where an AI summary informs a story, the underlying primary source is consulted directly. The standard of verification is the same one we apply to any tip: trust nothing until it is checked.

  1. Transparency and labelling

We disclose AI use to our audience when it is material — that is, when knowing about it would reasonably affect how a reader, viewer, or listener understands what we have produced.

As a guide:

  • Routine assistance that does not change the substance of the journalism — transcription, spell-checking, a suggested headline an editor approved — does not require a label, because a human remains fully responsible for the result.
  • AI-assisted content, where AI played a substantive role that a human directed and verified (for example, a translation produced with AI help), is disclosed where relevant.
  • AI-generated visuals or illustrations that appear in our coverage are clearly and visibly labelled as such, so they can never be mistaken for documentary images.

When in doubt, we disclose. The test is simple: would our audience feel misled if they later learned how this was made?

  1. Sources, confidentiality, and data protection

We do not enter confidential, personal, or sensitive information into external AI tools that could store it, share it, or use it to train their systems. This includes the identities and communications of sources, unpublished reporting, and the personal data of individuals.

Our handling of personal data complies with the Data Protection Act 2017 of Mauritius. Before adopting any AI tool, we consider where it stores and processes data, what it does with what we submit, and whether that is compatible with our legal and ethical obligations. Where a tool’s data practices are unclear or unacceptable, we do not use it for anything sensitive.

  1. Integrity of images, audio, and video

Photojournalism documents reality. We do not generate or alter news images, audio, or video with AI in any way that changes what actually happened, and we never present synthetic media as a record of a real event.

AI-generated illustrations may be used only where an illustration is clearly appropriate — for example, to depict a concept that cannot be photographed — and only when clearly labelled as an illustration. The line is absolute: genuine documentary material and synthetic material are never blurred.

  1. Bias, fairness, and language

We work across Kreol, French, and English, for a diverse Mauritian audience. AI tools can carry the biases of the data they were trained on, and can handle some languages — Kreol in particular — less reliably than others. We stay alert to this. AI-assisted language work is reviewed by people who understand the language and the community, and we do not allow a tool’s blind spots to distort how any group is represented.

  1. Copyright and intellectual property

We respect intellectual property. We do not present AI-generated material as original work where that would be misleading, and we are mindful that AI tools may reproduce content drawn from others’ work. Where AI contributes to a published piece, we ensure the result does not infringe the rights of third parties.

  1. Corrections and accountability

An error is an error, whatever its source. When a mistake involving AI reaches our audience, we correct it openly and promptly under our normal corrections process, and we examine how it happened so it does not recur. We never use AI involvement as an excuse; the responsibility remains ours.

  1. Governance, training, and oversight

Responsibility for this policy sits with [the Editor-in-Chief / a designated AI editorial lead]. Their role is to keep the policy current, advise the newsroom on difficult cases, assess new AI tools before they are adopted, and ensure staff are trained.

All staff and regular contributors are briefed on this policy and on the responsible use of AI. New tools are reviewed for accuracy, data protection, and editorial risk before they enter our workflow. Difficult or novel situations are escalated to the policy owner rather than resolved informally.

  1. Review of this policy

AI changes quickly, and so will this policy. We review it at least once a year, and sooner if technology, law, or our own experience requires it. The current version, its effective date, and a way to contact us with concerns are available to our audience.