Gains:
- Ability to understand plagiarism, copyright and the legal status of artificial intelligence output (who owns it, what is protected) at a basic level and distinguish risky use
- Understanding the need for transparency and publisher/reader disclosure and being able to position the use of artificial intelligence honestly
- Ability to establish a sustainable work balance by recognizing how over-dependence erodes creative skill and original voice
The most valuable capital for a writer is trust: the trust of the reader, the publisher and colleagues. It is built on trust, authenticity, honesty and responsibility. When artificial intelligence (AI) enters the writing process, these three values raise new questions: To whom does this text belong? Is it considered plagiarism? Should I tell the reader? And perhaps most insidious: What do I lose as a writer if I lean too heavily on AI? This final unit is the ethical backbone of the module. Technical skills are important; but without the principles that guide them, even the fastest writer can make the biggest mistake.
Risk of plagiarism and similarity
Plagiarism is presenting someone else's work or idea as your own without citing the source. AI poses two types of risks in this regard. The first is direct: the AI can produce an output that is very close to, or even almost identical to, a text in the training data; If you use this without realizing it, there is a risk of plagiarism. The second is indirect: the seemingly “original” text produced by AI may reflect too closely the structure, argument set, or character of an actually existing work.
The protection method is twofold: (1) not using the AI output as is but transforming it with your original contribution, (2) passing it through a similarity/plagiarism check before publication. The stronger your own contribution, the lower the risk of similarity.
Caution: It is wrong to think that "AI wrote it, therefore it is original". AI generates from patterns of existing texts; the output may look dangerously similar to another work. The responsibility lies with the author who published the final text; “AI did it” is not a defence.
Copyright: who owns it, what is protected
Copyright is the rights granted to the creator of an original work. The copyright status of AI output is still an evolving issue around the world and varies by country; general trends are:
- In many legal systems, copyright protection requires human creativity. An output produced entirely by AI, with no human contribution, may be considered non-protective in some countries.
- Works shaped by human creative contribution (selection, arrangement, transformation, original addition) find stronger protection grounds.
- There is also a risk of violating someone else's rights, as the AI output may resemble copyrighted works in the training data.
Practical and safe attitude in this uncertainty: use AI as an assistant, put your own unique creative contribution to the work, check the output for similarity. In this way, your protection basis is strengthened and the risk of violating someone else's rights is reduced. If you are doing contract work (publisher, client, agency), be sure to read the contract terms regarding the use of AI; Many contracts regulate this.
The table below compares risky and safe use:
Status
Risk level
safe attitude
Publishing AI output as is
high
Convert with original contribution
No similarity check
high
Check before publishing
Not reading the contract terms
high
Read the AI article
Upload sensitive/confidential text to open tool
high
Safe vehicle, don't share unnecessary
Using AI as an assistant, additively
low
sustainable method
Transparency: should I say?
Transparency means honestly reporting the use of AI to relevant parties. The rule is not uniform; depends on context:
- Academic and journalistic context: Statements are generally expected; Some institutions have strict policies. Learn the rules in advance.
- Publishing: Many publishers and platforms require disclosure of AI use; Read the contract.
- Customer work: Be honest with the customer; Talk about how you will use AI and how you will use it.
- Personal/creative work: Although not a requirement, honesty builds trust.
General principle: to be honest rather than hiding. Hiding the use of AI turns trust against you rather than in your favor when it is revealed.
Tip: Give yourself this test: "Would I be offended if I made it clear to my reader/customer what AI support I use?" If your answer is yes, it's time to review either your usage or your transparency. Honesty extinguishes a risk that may explode later.
Overdependence: silent erosion
Perhaps this is the most insidious risk, because it does not happen all at once, but slowly. Writing is a muscle; If it is not used, it will atrophy. A writer who delegates every sentence, every idea, every correction to AI will lose:
Creative muscle: The ability to solve a difficult scene on one's own, to wrestle with the blank page and emerge victorious. This struggle is what develops writing; If it is transferred, development stops.
Critical judgment: The ability to sense what is good, what is cliché, what phenomenon is questionable. The author who has everything approved by AI loses his own filter.
Authentic voice: The author, who constantly writes in the average tone of the AI, slowly loses his own quirks, his recognizable trace; texts become homogeneous.
The sustainable balance is to use AI outside the creative core: let the idea come from you, the voice from you, the decision from you; Let AI speed it up, organize it, offer alternatives, but not take over the core. A rule of thumb: always do at least one stage (usually the first creative draft or final reworking) yourself; This way your muscle memory stays alive.
Contribution test: How much of this work is mine?
There is a simple self-assessment that embodies the originality debate: the contribution test. After finishing a text, ask yourself four questions. (1) Who came the idea — the original premise, argument, or angle, from me or the AI? (2) Whose structure — did I make the partitioning and flow decisions? (3) Whose voice is it — do the sentences carry my tone or the average tone of the AI? (4) Whose judgment is it—did I decide what stays and what comes out, which facts are true? If you can say "I" to all four questions, the work is legitimately yours and the AI remains a tool. If you call most of the questions "AI", there is no authorship, but montage; Both ethical and legal grounds are weakened.
Size
If "mine"
If "AI's"
idea
Original premise/angle
It's commonplace, it doesn't belong to you
structure
conscious order
mold skeleton
sound
Recognizable style
homogeneous tone
Judiciary
Audited decisions
Unapproved transfer
Tip: Perform contribution testing at the end of every major project. If you see “AI” responses increasing over time, this is an early warning of overdependence; In the next project, consciously take at least one dimension back entirely to yourself.
Practical ethics templates
The following three templates translate ethical principles into daily practice. The first is a transform control that controls customizing the AI output:
Below is an AI sketch. REWRITING. Task: Evaluate how much of this text is "general" and how much of it contains an original angle or a specific observation. Mark the sections where the original contribution is weak; I will transform those places with my own ideas and voice. Just detect, don't stuff.
The second is to draft a transparency statement to a client or publisher:
Task: Draft the following modus operandi as a one-paragraph, honest and clear "AI usage statement":- The idea, structure and final voice are mine.- I used AI in the [draft / proofreading / brainstorming] stage.- I verified all facts from the primary source. Do not exaggerate or conceal; Use simple and reassuring language.
The third is a self-checklist to run before publication:
Task: Check the following text for preparation for publication and make a checklist: 1) are there any unverified facts/quotes, 2) are there passages that are too similar to another work, 3) is the tone homogeneous/unidentified, 4) is the original contribution sufficient? Mark each item as "ok / attention / missing". Leave the decision to me.
Tip: Make these three templates a “pre-publish routine.” Ethics is not a one-time decision, but a small set of habits that are repeated with each project; When you make it routine, the risks decrease on their own.
three mini cases
Case 1 — Similarity capture. A content writer checked an AI-produced episode for plagiarism before publishing; found that some of the text was 40 percent similar to an existing article. He rewrote and originalized it. Without the check, he could have faced a serious allegation of plagiarism.
Case 2 — Transparency gain. A freelance writer clearly explained to his client that he used AI to draft and proofread, but verified all the facts and wrote the text in his own voice. The customer was pleased with this honesty and the business relationship was strengthened. If he had hidden it, trust could have been broken when it was found out later.
Case 3 — Erosion awareness. A novelist found that after six months of having the AI write everything, he was having trouble writing a scene on his own; "My muscles are rusty," he said. He merely edited and brainstormed YZ, and began writing the first drafts himself again. Within a few weeks, his fluency and voice returned.
Weak approach / Strong approach
Weak approach:
Having AI write the entire book and publishing it under my name, without changing anything, without checking the source, and without telling anyone.
This approach carries all the risks of plagiarism, copyright, fact and trust.
Powerful approach:
The idea, sound and structure are from me; I use AI in drafting, alternating and editing. I verify every fact from the primary source. I transform the output with my own voice. I check for similarity/plagiarism before publication. I read the AI clause in the publisher/client agreement and act transparently. I always do at least one creative stage myself.
The difference: responsibility, originality and honesty belong to the author; AI accelerator but not core.
Common mistakes
- Thinking "AI wrote it, it is original". The output may resemble other work; Transformation and control are essential.
- Bypassing the similarity/plagiarism check. Any similarity revealed after publication carries a heavy price.
- Not reading contracts and corporate policies. The use of AI is regulated in most contracts.
- Hiding the use of AI. When exposed, trust turns against you; honesty protects.
- Delegating every stage to AI. The creative muscle, judgment, and voice atrophy over time; You become homogeneous.
In summary
AI ethics in writing stand on three pillars: originality, honesty, and responsibility. Transform and check the output with original contribution against the risk of plagiarism and similarity; put your own creative input in copyright uncertainty and read contracts; Honestly position AI use with transparency. To avoid the most insidious risk of over-dependence, always keep the creative core within yourself and always do at least one stage yourself. AI can augment your skill, but if it replaces it, it will begin to erase you. The balance is to use it as an assistant and keep the heart of the writing for yourself.
Application task
Review your own AI usage practice according to the "strong approach" items in this unit. Take a text of yours and: (1) evaluate the proportion of your original contribution, (2) put it through a similarity check, (3) apply the transparency test (“do you mind if I say it openly?”), (4) count how many creative steps you did yourself and how many with an AI in the last week. Write a personal “sustainable balance” rule: which step will you always do yourself?
checklist
- [ ] Have I transformed the AI output with my original contribution?
- [ ] Did I check for similarity/plagiarism before publication?
- [ ] Have I read the AI clause in the publisher/client agreement?
- [ ] Have I transparency tested my use of AI?
- [ ] Have I moderated safe tools and unnecessary sharing of sensitive/confidential text?
- [ ] Have I done at least one creative phase (draft or final revision) myself?
Module Exam
1. A novelist puts a chapter produced by artificial intelligence into his book without changing it, rewriting it in his own voice, and without verifying the facts. What is the fundamental authorial flaw in this approach?
- A) Publishing the output without crediting it and verifying its facts; Ignoring that creative responsibility lies with the author ✔
- B) Using artificial intelligence in fiction writing is prohibited under all circumstances
- C) Artificial intelligence always writes better dialogue than humans
- D) The section is not summarized in a table
Description: Artificial intelligence is a draft generator and assistant; The original voice, creative decision and final responsibility belong to the author. Publishing unverified and unauthenticated output as is risks both originality and the credibility of the text; A fabricated fact becomes a published error.
2. What is it called when artificial intelligence language models produce a source, quote or fact that does not actually exist in a fluent language as if it were true?
- A) Synopsis
- B) Hallucination ✔
- C) Proofreading
- D) Localization
Explanation: A hallucination is when the model presents non-existent information (fabricated source, fake quote, wrong date) in a safe and fluent sentence. In writing and editing, this is the most dangerous trap because fluency creates the impression of accuracy; the only protection is systematic verification linking to the primary source.
3. What is the best approach when requesting a content plan (outline) from artificial intelligence for a non-fiction article?
- A) Having the plan established according to the ready resources in the artificial intelligence's memory and not verifying it
- B) Having all subheadings adapted to artificial intelligence to keep the plan as long as possible
- C) Defining the target reader, main claim, length and tone and asking for a plan, then checking each item with its own knowledge ✔
- D) Saying 'write a plan about this' without giving any context and using the output as is
Explanation: A good output arises from a good context. Defining the target reader, main claim, length and tone, and requesting a content plan puts the AI at the service of your argument. Saying 'write a plan' without context produces a framework that is generic and not your own.
4. The AI produced a character sketch, but the character is full of clichés like 'mysterious, powerful and traumatized'. What is the best attitude as a writer?
- A) Even though it is a cliché, using the output as it is, because artificial intelligence knows better
- B) Completely delete the character and stop using artificial intelligence
- C) Repeating the same prompt over and over and waiting for a unique character to appear automatically
- D) Recognizing the cliché and reconstructing the character according to one's own observation with concrete, contradictory and original details ✔
Explanation: Artificial intelligence gravitates towards the most statistically probable, that is, the most generally accepted statements; This produces clichés. The author's job is to make this skeleton unique with his own observation, contradictions and specific details. Recognizing the cliché and replacing it with concrete, contradictory and original details makes the character alive.
5. A copy editor notices that a sentence correction suggested by AI subtly changes the meaning of the text. What is the correct behavior?
- A) Reject the correction that distorts the meaning or adapt it in a way that preserves the author's intention ✔
- B) Accepting the suggestion without caring about the meaning if it is grammatically correct
- C) Accepting all artificial intelligence suggestions collectively and without review
- D) Having the entire text rewritten by artificial intelligence
Explanation: In copy editing, even if grammar is correct, meaning and the author's intent must be preserved. AI sometimes shifts meaning for the sake of a 'smoother' sentence. The editor's job is to check each suggestion for meaning and sound, rejecting corrections that distort the meaning or adapting them to preserve the intent.
6. Which of the following is the most effective method that can be given to artificial intelligence to preserve the author's voice?
- A) Leaving the entire text in the default tone of the artificial intelligence
- B) Giving a voice guide of the writer's sample paragraphs, word choices, and tone notes and still reworking the output ✔
- C) Saying 'write like me' without giving any examples to artificial intelligence
- D) Choosing the most formal and general tone because it appeals to everyone
Description: When left to its own devices, AI produces a homogeneous, generic tone. Providing a 'voice guide' consisting of the author's word choice, sentence rhythm, tone and sample paragraphs brings the output closer to the author's style. However, the final stylistic decision and final reworking belongs to the author.
7. What is the most critical control when preparing the back cover text (blurb) of a book with artificial intelligence?
- A) Keeping Blurb as long as possible
- B) Explaining the end of the book clearly in the introduction
- C) Verifying that Blurb accurately reflects the book's actual content, tone, and balance of spoilers ✔
- D) Adding events that are not in the text to make them most striking
Description: The promotional text should reflect the reality of the book and not mislead the reader. AI may produce events that are intriguing but not in the text, or the wrong kind of promise. The author's task is to verify that the blurb is faithful to the text's actual content, tone, and balance of spoilers.
8. When receiving translation support with artificial intelligence, what is the most correct attitude when it comes to idioms and word games?
- A) Always translate idioms literally (word for word)
- B) Delivering the translation as it is without any comparison
- C) Completely delete word games and shorten the text
- D) Taking alternatives from artificial intelligence, making the cultural adaptation decision as a human translator and comparing the output with the source ✔
Explanation: Idioms, puns, and cultural references lose their meaning in literal translation. AI is useful as an outline and generator of alternatives, but the decision to culturally adapt (localize) requires the judgment of a human translator familiar with the target culture. By comparing the output with the source text, loss of meaning and tonal shift are checked.
9. Artificial intelligence for an article '2019, Journal of X, Yılmaz et al.' He cited a convincing source in the form. What's the first thing you should do as a writer?
- A) Finding the source in the primary source and verifying its existence and content; ✔ Do not use if not found
- B) Adding the source directly to the bibliography because it seems convincing
- C) Asking artificial intelligence 'is this source true' and trusting its answer
- D) Using the source several more times without checking it at all
Description: Artificial intelligence can make up sources, author names, and journal information that do not actually exist (hallucinate). Before including a source in the article, it is essential to find and verify it independently in the primary source (journal, database, official publication). Unverified sources are never used.
10. You want to find a plot hole in the plot with the support of artificial intelligence. Which method is the most efficient?
- A) Applying the first solution suggested by artificial intelligence without looking at the theme
- B) Give the event summary and rules and ask the artificial intelligence to list inconsistencies, motivation gaps and unsolved clues, and make the decision as the author ✔
- C) Just asking artificial intelligence 'is my novel good?'
- D) Ignoring the spaces because the reader won't notice
Description: Artificial intelligence is a powerful auditor in analyzing the plot in terms of cause-effect chain and character motivation. Giving him the summary of the events and the rules and saying 'list inconsistencies, motivational gaps, and unresolved clues' makes the overlooked gaps visible. However, it is up to the author to decide which solution suits the theme.
11. What risk should be considered before uploading a sensitive or unpublished manuscript to a publicly available AI tool?
- A) Uploading text improves translation quality
- B) There is no risk because all tools are the same
- C) The text can be stored or used in model training; ✔ poses a privacy and copyright risk
- D) Artificial intelligence always publishes the text automatically
Disclosure: An unpublished manuscript is the intellectual property of the author. Text uploaded to a public tool can be stored or used in model training, depending on the conditions of the tool; This creates privacy and copyright risks. It is necessary to choose corporate tools that have data processing contracts, that do not use your data in education, and to avoid unnecessary sharing.
12. Which is the most accurate statement about the copyright and originality status of the artificial intelligence output?
- A) Artificial intelligence output is always and automatically the original copyrighted work of the author
- B) Artificial intelligence output can never contain plagiarism
- C) Copyright does not concern authorship at all
- D) The output may resemble other works; The author must transform with his original contribution, check the similarity and take responsibility ✔
Description: Artificial intelligence produces patterns of texts in the training data; the output may be too similar to other works or may be preservation-ambiguous. The author must transform the output with his or her original contribution, control the risk of plagiarism and similarity, and be transparent with respect to publisher expectations. The responsibility lies with the author who published the final text.
13. What is the main risk of over-dependence (relying too much on AI) for a writer?
- A) Atrophy of creative skill, original voice and critical judgment; homogenization of text ✔
- B) Extremely slow writing speed
- C) Artificial intelligence refuses to work
- D) Too much originalization of the text
Explanation: Creative skill, fact-checking reflex and original voice only develop with use. Delegating every sentence, every idea, and every correction to artificial intelligence dulls the writer's muscle memory, style, and critical judgment, making the text homogeneous and generic. The sustainable balance is to use AI as an assistant and keep the creative core with the writer.
14. An editor wants to check the consistency of the style guide in a text in which the author has the style, using artificial intelligence. Which is the most correct approach?
- A) Completely ignoring the style guide
- B) Clearly state the rules of the style guide and have the text proofread, then review each suggestion for meaning and conscious choices ✔
- C) Saying 'fix' to artificial intelligence without giving the style guide
- D) Accepting all suggestions without examining them
Description: Style guide; It ensures consistent application of rules such as spelling, uppercase and lowercase letters, numbers and punctuation. Giving the AI the rules of the guide explicitly and having it check the text according to these rules increases consistency; However, each suggestion should be reviewed in terms of meaning and sound, and conscious preferences that do not comply with the rules should be preserved.