Editorial Policy

Last updated: October 20, 2026 

This page explains how thehype produces and checks what it publishes: on thehype.news, on our social accounts, on thehype radio and in the ai morning podcast. Please read it together with our Terms, Privacy Policy and Cookie Policy.

1. Who We Are

thehype is an independent media outlet covering artificial intelligence: new models, agents, creative tools, funding, and the decisions of major technology companies. It is operated by Crezee FZC, a company registered in the United Arab Emirates. Full company details are available in our Terms.

Our website's editorial decisions are made by our editor-in-chief and editorial team. Learn more about our mission and the story behind thehype on our About page. The people responsible for our journalism are listed on our Team page, and our authors are listed on our Authors page. Articles and other original editorial content published on the website are produced by human authors, who are identified on their work.

Our audio and broadcast formats, including thehype radio and ai morning, are produced through fully automated workflows powered by specialized AI editors. These editors are configured with detailed instructions and rules for selecting stories, evaluating sources, verifying claims, and preparing scripts for publication. Each story goes through the automated checks applicable to the workflow that produces it. Our editorial team develops and refines these workflows, but individual stories, scripts, and episodes are produced and published without human review before publication.

2. What We Cover

On thehype.news, we cover artificial intelligence across the following categories: news, AI agents, big tech, creative AI, funding, guides, and models. These categories reflect our current coverage and may evolve as the field develops.

We prioritize news that can be verified, including new models and infrastructure, product launches, funding rounds, decisions by major technology companies, and changes to developer tools and platforms. We favor primary sources, such as official announcements, documentation, regulatory filings, and code repositories, and distinguish confirmed developments from rumors and unverified reports. Unconfirmed reports are clearly labeled as such (see Section 6).

We publish content across social media and digital platforms, including X, LinkedIn, Threads, Bluesky, YouTube, Instagram, and others. Our platform presence evolves over time as we join new channels and develop new ways to reach our audience. We publish both content adapted from our website and formats created specifically for each platform, including:

• Comparisons: Side-by-side tests of AI models using our own testing methodology.
• Roundups: Data-driven rankings and weekly lists, including funding rounds, fast-growing GitHub repositories, model usage on OpenRouter, and popular open models. Some roundups are also published as animated rankings.
• Digests: Summaries of the day's or week's developments in AI, published as text posts, graphics, or short videos.
• Opinion and commentary: Our analysis of news, products, and statements by companies or individuals.
• Live event coverage: Coverage of conferences and major product announcements as they happen, sometimes followed by recap videos and written summaries.
• Breaking news and model releases: Short reports on significant announcements, with a clear editorial focus.
• Rumors: Reports based on information that has not been officially confirmed, including expected developments, clearly labeled as unconfirmed (see Section 6).
• Reviews, explainers, and prompts: Short videos featuring product reviews, explanations of AI concepts, or prompts and their outputs. Reviews and explainers may feature our mascot, Hypey.
• Interviews: Full conversations and highlights from our thehype talks series.

We also operate thehype radio (radio.thehype.news), a 24/7 AI-powered broadcast, and ai morning, a daily AI-generated news podcast. Both are produced through automated AI editorial workflows, as described in Sections 1 and 5.

3. How We Choose What to Cover

Our news discovery process is powered by a network of specialized AI editors developed and refined over an extended period. These AI editors continuously monitor a wide range of sources, collect relevant information, identify potential stories, and analyze developments across the AI industry. They work together through coordinated editorial workflows, following detailed instructions for source evaluation, fact-checking, and news preparation.

The system serves two purposes. It powers the automated production of news for our radio and other AI-driven formats, and it provides our human editorial team with a continuous flow of news and source material to inform coverage on our website. Editors and authors use this material alongside their own research and independent news discovery to select stories for articles.

Our sources include company blogs, official documentation, official company and individual accounts where authenticity can be established, public repositories, model catalogs, video channels, and other relevant sources. When evaluating a potential story, we consider the following:

• Verifiability: Are the key claims supported by evidence, and can important facts be checked against reliable sources?
• Source quality: Does the information come from a primary source where available, or can it be corroborated by other reliable sources?
• Significance: Does the development represent a meaningful change, announcement, result, or decision rather than merely a promotional claim?
• Relevance: Is the story useful or interesting to readers who follow, use, or build with AI?
• Accuracy and timing: Are names, figures, dates, and timelines consistent with the available evidence? Is the information current, and has anything changed since the original announcement?
• Additional verification: Where appropriate, can we examine the underlying documentation, inspect a public repository, or test the product ourselves?

We do not treat attention or promotional language alone as sufficient grounds for coverage. Claims should be supported by evidence appropriate to the story, and unconfirmed information is clearly identified as such (see Section 6).

4. Sourcing and Fact-Checking

Our coverage is based on verifiable information and sources that readers can consult wherever possible. We distinguish confirmed facts from claims, opinions, and information that remains unverified.

Sources and Evidence

We support claims with primary sources, such as official announcements, documentation, research papers, public repositories, regulatory filings, and pricing pages. We may also use attributed reporting from other publications and our own tests. When a claim comes from another source, we attribute it appropriately.

Checking Information

We prioritize primary sources where available and do not rely solely on summaries from other media outlets, aggregators, or AI tools when the underlying source can be consulted. We check claims against the available evidence and consider whether information is corroborated by other reliable sources. Where relevant, we examine documentation, inspect repositories, or test products ourselves.

We distinguish official company statements from information reported by third parties. An official statement establishes what a company has said, but does not automatically establish that every claim within it is independently verified. Information that cannot be confirmed may be covered when editorially justified, provided its status is clearly explained (see Section 6).

Numbers, Quotes, and Product Details

We use figures and details as reported or measured, without presenting estimates as confirmed results. When we make our own estimates, we identify them as such. Quotes, product names, version numbers, prices, and benchmark results are checked against the relevant original statement or source wherever possible. We provide context where needed to avoid misleading comparisons or interpretations.

Links and Attribution

We link to the sources used in our reporting wherever applicable. When an article relies on external sources, we identify and link to them so readers can consult the original information. Sources may be linked directly in the article or listed at the end, depending on the format. We clearly distinguish our own reporting and analysis from information, claims, and opinions attributed to others.

These linking and attribution practices apply to our editorial articles and other content prepared by our team. In our automated broadcasts, single-story radio segments include a clickable link to the source. Compilations, which make up most radio segments and all ai morning episodes, do not list sources for individual stories.

Named and Anonymous Sources

We identify sources by name where appropriate and possible. Information from an anonymous source may be used when the source's identity is known to us or their credibility can otherwise be reasonably assessed, and the information is relevant to the story. We make clear to readers when information comes from an undisclosed source. Anonymous information is assessed carefully and corroborated where possible.

Source Protection

We protect the identity of confidential sources and do not disclose it without their explicit consent, except where disclosure is required by law.

Headlines

Our headlines accurately reflect the content of the story. We do not deliberately exaggerate claims, promise information that the article does not deliver, or conceal the central fact simply to attract clicks.

Fabricated Sources and Quotes

We do not knowingly publish fabricated sources, invented quotations presented as genuine, or citations that do not support the claims attributed to them.

Images and Visual Materials

We use images, screenshots, illustrations, and graphics to support our coverage. These materials must not misrepresent the facts or create a misleading impression.

We may use images and visual materials from company announcements and press materials released for media use, from presentations and public events, and from licensed or credited sources. Where appropriate, we identify the source or creator of externally sourced materials. We do not add separate source credits to our own original photographs, videos, or other materials created by our team.

Our illustrations and graphics are made with AI tools, as described in Section 5.

5. Use of AI

Articles, Posts and Videos

Our AI-powered news system monitors sources, gathers news, and helps our editorial team identify relevant developments. Authors use these sources and materials in their research and may also discover and investigate stories independently.

Our human editorial team decides what articles, social media posts, and editorial videos to publish. AI tools may assist with research, translation, outlining, drafting, and editing, but authors and editors review and approve the content they create before publication, including checking the accuracy of its claims.

This applies to editorial content prepared by our team, not to automated broadcasts such as thehype radio, which follow the separate workflow described below.

We do not enter information that could identify a confidential source into AI tools.

Illustrations and Graphics

We use AI tools to create illustrations and graphics for our editorial content. These visuals may be selected, directed, edited, or adapted by our team. We do not present AI-generated illustrations as photographs of real events or use visuals intended to mislead readers about what actually happened.

Comparisons

Our AI model comparisons use tests designed for the specific task being evaluated. The methodology, conditions, and other relevant details are provided in each test's accompanying content. Pricing information is sourced from the provider's published pricing, OpenRouter, or Artificial Analysis with the source identified in the post. Outputs shown in our comparison videos are actual outputs generated by the models being tested.

thehype radio

thehype radio is a fully automated, 24/7 AI-powered broadcast. Its hosts are AI-generated voices, identified as AI hosts on the radio and on our social media accounts.

Each news category has its own AI editor. These AI editors monitor sources, select stories according to their own editorial logic, write scripts, and check them against their sources before converting them to speech. Scripts that fail these checks are not aired.

No person reviews individual scripts before broadcast. Our editorial team regularly listens to the broadcast, updates the sources, and systematically improves the AI editors and the underlying models. Single-story segments name their source with a clickable link. Most segments are compilations of news from the past half hour or hour, and these do not list sources for individual stories. It is distributed in audio and video formats through platforms including YouTube, X, and TuneIn. On YouTube, only the current stream is public, and earlier recordings are kept unlisted. On X, recordings of broadcasts remain available. TuneIn carries the live broadcast only, without recordings.

ai morning

ai morning is a daily news podcast created and published automatically using AI-generated voices. Each episode covers selected top news stories from the previous day. Each episode is a compilation of several stories, so episodes do not list sources for individual stories.

The production process follows the same approach as thehype radio: AI editors select stories, prepare scripts, and check them against their sources before the scripts are converted to speech. Stories are selected according to the sourcing and verification principles described in Sections 3 and 4.

Episodes are published on platforms including YouTube, Spotify, Apple Podcasts, and TuneIn.

Hypey

Hypey is our AI-created brand mascot, not a real person. Hypey does not have personal opinions or independent views. Videos featuring Hypey are written and produced by our editorial team.

6. Rumors and Unconfirmed Information

We clearly label rumors as such, in the headline or opening sentence, or at the start of the item in audio and video formats, and identify the source of the information, or say that it comes from an undisclosed source. We maintain a list of sources we consider reliable, but may also assess and use new sources when reporting rumors. We consider each source's track record and a company's observed release patterns when evaluating claims about potential announcements or releases.

We do not present predictions or unconfirmed claims as established facts. When a rumor is subsequently confirmed, denied, or materially clarified, we may update the relevant article or post where editorially warranted.

7. Interviews

Our interviews are prepared and conducted by our editorial team. Guests are invited by us. Where appropriate, we may share an edited interview with a guest to check factual accuracy and context before publication.

Guests may point out factual errors or provide clarification, but they do not control our questions, editorial decisions, or the final content of the interview.

Any financial relationship or compensation connected to an interview or its publication is disclosed where relevant, so readers can distinguish editorial interviews from paid or sponsored content.

8. Independence, Conflicts of Interest and Sponsored Content

Editorial Independence

On our website, editorial decisions are made by the editorial team. In our automated formats, they are made by AI editors configured by the team (see Sections 1 and 5). Both are guided by newsworthiness, relevance, and reader interest. Commercial relationships do not determine what we cover or what we say about it. Clients, sponsors, and partners may review and provide feedback on content created as part of commercial collaborations, but independent editorial decisions remain the responsibility of the editorial team.

Commercial Relationships and Sponsored Content

We work with commercial partners on sponsored content, media partnerships, events, and other collaborative initiatives. Content produced as part of a paid collaboration is clearly identified as sponsored, partner, or promotional content, as appropriate. Commercial arrangements may include event attendance, journalist travel, or other production-related expenses. Where these arrangements are relevant to published coverage, we disclose them as appropriate and distinguish sponsored content from independent editorial work.

Conflicts of Interest

Members of the editorial team are expected to disclose any actual or potential conflict of interest to the editor-in-chief. This includes financial interests, investments, advisory roles, paid relationships, or other affiliations that could reasonably be perceived as influencing their work. The editor-in-chief reviews potential conflicts and determines whether disclosure or other measures are necessary. Where relevant, conflicts are disclosed to readers.

9. Corrections and Updates

Updates

We update articles when significant new information becomes available or when there is a specific editorial reason to revisit a story. We do not routinely review our entire archive for updates. Instead, previously published content is reviewed on a case-by-case basis in response to new developments, verified information, reader feedback, or other relevant circumstances.

• Updated articles display a "Last updated" timestamp while retaining their original publication date.
• Significant updates include an editor's note explaining what changed.

To suggest an update, please contact us at editorial@thehype.news with the article URL and the specific information or claim that may need revisiting.

Corrections

If we publish inaccurate information, we correct it as soon as reasonably possible after becoming aware of the error. The nature of the correction depends on the severity of the error:

1. Minor errors (such as typos, formatting issues, or incorrect links) are corrected directly without a formal correction note.
2. Material errors (such as inaccurate facts, figures, or statements that affect the meaning or conclusions of a piece) are corrected, with a dated editor's note explaining the change. The article's "Last updated" timestamp is revised accordingly.
3. Retractions are reserved for cases where a piece is fundamentally unreliable or materially wrong and cannot be adequately addressed through a correction or update. Where a piece is retracted, we provide an explanation where appropriate.

On social platforms, where editing options may be limited, we publish a follow-up post or a pinned correction, as appropriate.

On thehype radio, segments are not repeated. Each script is checked against its sources before it is aired, and a script that fails the check is not aired. If an error reaches the air anyway, the broadcast itself cannot be corrected. We then fix the cause: we add or tighten verification rules and remove unreliable sources from the system. If a recording of the broadcast remains available on X, the rule below on published recordings applies to it.

Where a material error affects an already published audio or video recording, we assess whether the recording should be corrected, replaced, or removed, depending on the nature of the error and the options available on the relevant platform.

We remove published content only in exceptional circumstances, such as when legally required, when it contains sensitive information that cannot be safely redacted, or when it is substantially inaccurate and cannot be adequately corrected or updated.

To report an error, please contact us at editorial@thehype.news with the article URL and the specific claim in question. We aim to review correction requests within 48 hours of receipt. If an error is verified, we aim to correct it as soon as reasonably possible. More complex cases may require additional time.

10. Feedback and Contact

If you have questions about our editorial standards, concerns about our coverage, or feedback on our content, please contact us.

You can reach out to us to:

• Report factual errors or request a correction or update.
• Flag potentially misleading, inaccurate, or outdated information.
• Raise concerns about our reporting, sourcing, or use of AI-generated content.
• Ask questions about our editorial standards or how we make editorial decisions.
• Share feedback or suggest improvements to our coverage.
• Submit a story idea or suggest a topic for future coverage.

For corrections and update requests, please include the article URL and the specific claim or information in question. We take substantive reader feedback seriously and review correction and update requests in accordance with the process described in Section 9.

11. Changes to This Policy

We update this page as our work changes. The date at the top shows the last update.

This policy describes our editorial practice. It is not a promise or a guarantee, and it does not create contractual obligations beyond those in our Terms.