AI Angst is a publication about artificial intelligence. It would be inconsistent, and dishonest, not to tell you how AI is used in making it.
This page explains what role artificial intelligence plays in our content, what human oversight is applied at every stage, what that means for the accuracy of what you read, and what we will never use AI to do. We believe transparency about AI use in publishing is not a legal formality. It is an ethical baseline, and one we hold ourselves to precisely because we cover this subject.
1. What We Use AI For
AI Angst uses large language models (LLMs), including Claude by Anthropic, as editorial tools in the production of content. This includes:
Research assistance: Summarising source documents, extracting key facts from lengthy reports, and cross-referencing information across multiple primary sources.
Drafting and structuring: Generating first drafts of articles, section outlines, FAQ content, and structured data (JSON-LD schema) based on detailed editorial briefs and verified source material supplied by our editorial team.
Formatting and HTML production: Generating publication-ready HTML pages from editorial templates and approved content specifications.
Translation assistance: Supporting the processing of source materials in languages other than English, including German-language research documents, to inform English-language editorial output.
SEO and metadata: Generating meta descriptions, schema markup, keyword structures, and structured FAQ content for search and discovery optimisation.
Plain language summary: AI helps us write faster and more consistently. It does not replace the editorial judgment, source verification, fact-checking, or final approval that every piece of content on this site goes through before publication. AI is a tool. The editorial decisions are human.
2. What We Do Not Use AI For
The following are human-only editorial responsibilities at AI Angst. AI tools are not used for, and do not replace:
Source selection and verification: Every primary source cited on this site — research papers, press statements, news reports, academic studies, quotes, is selected, read, and verified by a human editor. AI does not choose what we cite.
Factual accuracy decisions: When a fact is in dispute, ambiguous, or unverified across sources, a human editor makes the final determination of what we publish and how we frame it.
Editorial position and opinion: The perspective, analysis, and critical framing in AI Angst articles reflect the views of our human editorial team. AI does not form opinions on our behalf.
Quoting real people: All quotes attributed to named individuals on this site are sourced from verified primary or secondary reporting. AI does not generate, paraphrase, or reconstruct quotes.
Final publication approval: No content is published without human review and approval. AI-generated drafts are always reviewed, edited, and signed off by a member of the editorial team before they appear on the site.
3. Human Editorial Oversight: How It Works
Our content production process follows this sequence for every article published on AI Angst:
| Stage | Who Is Responsible | What Happens |
|---|---|---|
| Topic selection | Human editor | Editorial team identifies the topic, angle, and publication rationale |
| Source research | Human editor + AI assistance | Primary sources are located, read, and verified by the editor; AI helps process and summarise long documents |
| Editorial brief | Human editor | Detailed prompt specifying all facts, quotes, sources, structure, and content requirements, AI receives only what the editor has verified |
| First draft | AI tool (LLM) | Draft article generated from the editorial brief and verified source materials |
| Editorial review | Human editor | Draft reviewed line by line for factual accuracy, tone, attribution, and compliance with editorial standards |
| Fact-checking | Human editor | All statistics, quotes, and factual claims cross-referenced against original sources |
| Final approval | Human editor | Signed off by editor before publication. No AI-generated content is published without this step. |
The publisher and editor-in-chief of AI Angst is Jans Bock-Schroeder. All content published on this site is ultimately his editorial responsibility.
4. Accuracy, Errors, and Corrections
AI language models can and do produce errors. Known failure modes include:
Hallucination: AI systems may generate plausible-sounding but false statistics, quotes, or citations. Our fact-checking process is specifically designed to catch these, every verifiable fact in our articles is checked against a named source before publication.
Outdated information: AI models have knowledge cutoffs and may produce information that was accurate at training time but is no longer current. We address this through live web research and by citing source dates explicitly.
Paraphrasing errors: AI may subtly misrepresent source material when summarising. Human editorial review checks that the meaning of primary sources is accurately represented in our articles.
Despite these processes, errors may occur. If you identify a factual inaccuracy, an incorrect attribution, or a misleading statement in any AI Angst article, please contact us immediately. We take corrections seriously and will update or correct published content when an error is confirmed.
5. Image and Visual Content
Images used on AI Angst may be sourced from the following:
Licensed photography: Images obtained under appropriate commercial licences from photography services.
Press and promotional materials: Images supplied by organisations for press use, credited where required.
AI-generated imagery: Some images on this site may be generated using AI image tools. Where this is the case, the image is illustrative and does not depict a real person, location, or event unless explicitly stated.
AI Angst does not use AI-generated imagery to impersonate, misrepresent, or deceive readers about the nature of depicted subjects. We do not generate fake photographs of real people or use synthetic images to illustrate events as if they were photographic documentation.
6. Why We Disclose This
We are not required by law to disclose AI use in our editorial content at this time. We disclose it because we believe you have the right to know how the content you read is made.
AI Angst covers the impact of artificial intelligence on society, culture, and truth. A publication that does not apply its own editorial standards to the use of AI in its own production would be, to use the word we use for that kind of content, slop.
We are not slop. We use AI as a tool under human editorial direction, with human fact-checking, human sign-off, and human accountability for what we publish.
If that ever changes, if our process becomes less rigorous, if AI use expands beyond what is described here, or if industry standards or legal requirements evolve, we will update this page to reflect it. The date at the top of this page shows when it was last revised.
7. Regulatory Context
The regulatory landscape around AI content disclosure is evolving rapidly. The following frameworks are relevant to AI Angst's operations:
EU AI Act (2024): The European Union's AI Act introduces requirements around transparency for certain AI-generated content, including obligations on providers of general-purpose AI systems. As these provisions come into force, AI Angst will review and update its practices accordingly.
EU Digital Services Act (DSA): Requires transparency around content moderation and algorithmic systems for platforms meeting certain scale thresholds. We monitor developments as they apply to independent publishers.
FTC Guidelines (US): The US Federal Trade Commission has published guidance on AI-generated endorsements, testimonials, and content. AI Angst does not use AI to generate fake testimonials, reviews, or endorsements.
Google Publisher Guidelines: Google's helpful content guidance and spam policies apply to all content indexed by Google Search. AI Angst's editorial process is designed to meet the "people-first" content standard: content produced for a human audience, under human editorial control, with genuine expertise and accountability behind it.
8. Questions and Contact
If you have questions about how AI is used in AI Angst content, how a specific article was produced, or how to report a suspected error or inaccuracy, please contact us:
AI Angst
Publisher: Jans Bock-Schroeder
28 Avenue Foch, 75116 Paris, France
Email: mail@aiangst.com
Contact form: aiangst.com/contact
About AI Angst
Who we are, what we believe, and why we started this publication.