Gains:
- Ability to design the entire production line from idea to delivery as an artificial intelligence-supported workflow and plan the tool stack
- Ability to link quality control, versioning, file/copyright registration and delivery steps to a checklist
- Ability to institutionalize a governance framework where artificial intelligence accelerates, but creative direction, authenticity and publication approval remain in the hands of humans
Module Exam
1. An editor delivers the rough cut produced by artificial intelligence directly to the customer without any monitoring or correction. What is the fundamental mistake in this approach?
- A) Delivering the artificial intelligence output without monitoring, narrative and continuity control; Ignoring that the responsibility for the final cut lies with the human being ✔
- B) It is strictly forbidden to use artificial intelligence in rough fiction
- C) Artificial intelligence always makes the video appear longer than it is
- D) The fiction is not presented on a timeline
Description: AI can statistically/automatically select the best moments, but may miss narrative flow, emotional transition, and continuity errors. Responsibility for the final cut and the version to be published belongs to the editor; unverified output is like an unverified fiction.
2. What is the best approach when uploading raw footage of an unreleased client project into a publicly available AI tool?
- A) Uploading all raw footage and customer name as is for speed
- B) Just change the filename and upload the content as is
- C) Obtaining verbal approval from the customer is sufficient in all cases.
- D) Using contracted/corporate tools that do not use your data in education, observing confidentiality and copyright, and not sharing unnecessary sensitive material ✔
Disclosure: Raw footage and client material are generally covered by a non-disclosure agreement (NDA) and copyright. Unpublished content should not go to an external server; If possible, contracted corporate tools that do not use your data in model training should be preferred and sensitive material should not be shared unless necessary.
3. What is the most critical thing to do before cloning a person's voice with artificial intelligence and using it in an advertisement?
- A) Checking whether the voice sounds natural enough
- B) Comparing whether cloning is done with the cheapest tool
- C) Obtaining clear, written consent with a specific scope from the owner of the voice ✔
- D) Changing the voice beyond recognition, making consent unnecessary
Explanation: A person's voice is a personal property and in most cases copyright/contractual. For the use of voice cloning, clear, written consent must be obtained from that person with a specific scope (where, for how long, for what purpose); Voice imitation without consent is an ethical and legal violation.
4. What must the editor do before publishing the automatic transcription output as subtitles?
- A) Verify proper names, terms, punctuation and time codes by comparing the text with the original audio ✔
- B) Exporting the transcript directly as SRT without reading it at all
- C) Just change the font and publish
- D) Duplicating subtitles with automatic translation and not checking any of them
Description: Automatic transcription frequently misspelles proper names, technical terms, abbreviations, and punctuation; mixes similar sounds. Time codes may drift. The editor should compare the text with the original audio and verify terms, names, and reading speed.
5. Which of the following is a real risk for a clip created with a text-to-video generator?
- A) The file size of the clip is always too small
- B) Frame-to-frame inconsistency, artificial detail and resemblance to copyrighted style; ✔ by labeling
- C) Productive video can never contain audio
- D) The clip can only be produced in black and white
Description: Generative video runs the risk of frame-to-frame inconsistency, artificial/unrealistic details, and unknowingly resembling a copyrighted style/character. Additionally, many platforms require productive content to be tagged. Checking and tagging required before publication.
6. What is the best approach when using a scenario draft produced by artificial intelligence?
- A) Shooting the draft without changing it, because artificial intelligence produces the most original idea
- B) To throw away the scenario completely, because artificial intelligence is useless
- C) Only change the character names and use them as they are
- D) Accept the draft as a beginning, eliminate clichés, and revise it according to brand voice and originality ✔
Description: The AI scenario can often be cliché, generic and lacking in brand voice; It may suggest unreal scenes. An outline is a starting point; Creative direction, originality and brand harmony must pass through the hands of the screenwriter.
7. What is the ethical limit when using sound cleaning and image enhancement with artificial intelligence in a documentary interview?
- A) Reducing noise and clarifying is allowed, but adding non-existent detail and changing reality/meaning is unethical ✔
- B) No artificial intelligence tools can be used in the documentary
- C) It is free to change the meaning of the content if it will be more interesting.
- D) Any intervention is acceptable as long as the image quality increases
Description: Enhancement tools can reduce noise and increase clarity; However, it should not distort reality by adding non-existent details (fabricated facial detail, changed words). Intervention that changes the meaning and reality of the content in news and documentaries is unethical.
8. What is the main mistake to avoid when generating headlines and thumbnails with artificial intelligence?
- A) Keeping the title short
- B) Using text in thumbnail
- C) Producing misleading clickbait headlines and images that contain claims that are not in the content ✔
- D) Producing and testing multiple title options
Description: Artificial intelligence may suggest clickbait titles/images that contain claims that are not in the content to increase views. This undermines audience trust and platform rules. The title and image should honestly represent the content.
9. What is the best approach in terms of copyright when using generative music or stock images?
- A) Assuming no license checks are required because it produces AI
- B) Verifying the vehicle's license, commercial use rights and attribution/registration requirements before use ✔
- C) Using every music found on the internet by citing it as a source
- D) Considering copyright only for music and considering visuals as free
Explanation: Each vehicle's license, commercial use permit and attribution requirements are different. Just because a piece of content is 'AI-generated' does not automatically make it royalty-free. Before use, license terms, commercial use rights and attribution/registration obligations, if any, must be verified.
10. Which of the following expresses the correct limit of the use of artificial intelligence in video production?
- A) Artificial intelligence can make all creative and publishing decisions automatically
- B) Artificial intelligence can only be used to write subtitles and not at any other stage
- C) Artificial intelligence is an assistant and accelerator; Responsibility for creative direction and publication approval lies with the human being ✔
- D) Artificial intelligence can publish content directly without human approval
Description: Artificial intelligence assistant, draft generator and accelerator. Responsibility and approval for the creative direction, brand voice, authenticity and final cut to be released always rest with the competent professional.
11. What is the most common mistake when automatically extracting short clips from a long video with artificial intelligence?
- A) Producing clips in vertical format
- B) Adding subtitles to clips
- C) Broadcasting moments out of context, cut off mid-sentence, or uncontrolled in a way that can be misunderstood ✔
- D) Producing multiple clip options
Description: The automatic clip extractor can select a moment out of context; may cut off in the middle of the sentence or highlight a section that may be misunderstood. Clips should be checked for context, cut-off point and semantic integrity before broadcast.
12. What is the correct use of images produced by artificial intelligence for storyboard?
- A) Using visuals as a planning/prevision tool that expresses the intention and shooting the actual footage according to physical and budgetary constraints ✔
- B) Substituting the produced image for the finished shot and deeming the shot unnecessary
- C) Delivering the images to the customer as a finished work
- D) Copyright should not be considered in visual production
Description: The images produced are a planning and communication tool (previz); Explains the intention to the team. Actual shooting is subject to budget, location, actor and physical constraints. Additionally, the image produced may resemble a copyrighted style; The team determines the final visual direction.
13. What is the right attitude towards AI translation and dubbing when producing multilingual dubbing?
- A) Publishing the translation and dubbing without verification by someone who knows the target language
- B) A person who knows the target language verifies translation, terminology, tone and synchrony and maintains meaning/brand integrity ✔
- C) Assuming that AI translation is always perfect
- D) Approving dubbing based on sound quality only
Description: AI translation may misrepresent idioms, cultural references and terms; lip sync and intonation may be unnatural. A person who knows the target language must verify the translation and dubbing and maintain the integrity of the brand and meaning.
14. What is the best practice when using a realistic human image (like a deepfake) produced by artificial intelligence in a published video?
- A) Hiding the source and method of production as long as it is realistic
- B) Not specifying any tagging as it will reduce the views.
- C) Leaving the responsibility entirely to the artificial intelligence tool
- D) Obtaining the necessary consent, transparently stating that the content is produced by artificial intelligence and complying with the platform labeling rules ✔
Description: Generating images that resemble a real person requires consent from the person; In order not to be misleading, it should be transparently stated to the audience that the content is produced by artificial intelligence and platform labeling rules should be applied. The responsibility lies with the producer.