Big decision on AI Copyright, What we know of Midjourney V7, and OpenAI's response to DeepSeek
All the news from the AI World for creatives
Good day, creatives
In today’s Newsletter:
1️⃣ What the U.S. Copyright Office’s decision means for creatives
2️⃣ Midjourney’s V7: Video, 3D, and everything we know
3️⃣ OpenAI’s response to DeepSeek is to launch o3-Mini
4️⃣ Quick news you may have missed
5️⃣ Prompt Fight: Golden Stars
6️⃣ Video AI of the week
Exciting News: A new name, a clearer focus
I'm rebranding to The AI Video Creator. When I first launched this newsletter, AI Video felt like a distant possibility. But today, it's a growing reality — and it's where my focus has naturally evolved.
With Midjourney's upcoming video features (see below), this change reflects my commitment to exploring and sharing everything about AI-driven video creation.
Let’s go.
AI and Copyright: What the U.S. Copyright Office’s decision means for creatives
In a landmark update that carries significant implications for creative industries, the U.S. Copyright Office has clarified how copyright law applies to works generated with the assistance of artificial intelligence (AI).
This development comes as part of the Office’s second report, titled “Copyright and Artificial Intelligence, Part 2: Copyrightability.”
While it doesn't resolve every question about AI and intellectual property, it provides much-needed clarity on how creators can protect their work when incorporating AI tools in the U.S, and this decision will have an impact in other parts of the world.
The Verdict: AI is a Tool, not an Author
After reviewing over 10,000 public comments and conducting nearly a year of research, the Copyright Office concluded that AI can assist in creating copyrightable works, but human input is crucial.
Merely typing prompts into an AI system isn’t enough to claim ownership.
Why? Because it might interpret your input differently than expected (we all know this is true), making prompts themselves insufficient for establishing creative authorship.
The Copyright Office reaffirmed a long-standing principle: copyright applies to human creativity, not machine-generated outputs.
What’s not Copyrightable?
Certain uses of AI tools are not eligible for copyright:
Raw outputs from tools like ChatGPT or Midjourney, even if prompted with great care.
Raw single Video clips generated by AI platforms like Runway, Pika, etc.
As an example, asking ChatGPT to “write a sci-fi story” and submitting the unedited result for copyright would be akin to claiming authorship simply for suggesting an idea.
Simply put, don’t be lazy.
Creation of greater works of art is still hard. AI just opens more possibilities.
What might be copyrightable?
If a human demonstrates substantial creative input, AI-generated works may qualify for copyright protection:
Heavily editing an AI-generated story.
Integrating AI elements within a larger, original work.
Editing and arranging multiple AI video clips to form a coherent film.
Films using AI-generated special effects, for example, can still be protected as a whole.
Why is this important?
This decision offers reassurance for businesses and creators concerned about protecting their intellectual property. Brands that use AI-generated media can retain protections for distinctive creative elements. For instance, Coca-Cola’s controversial AI-generated holiday commercial could maintain copyright as long as human creative decisions were pivotal.
Midjourney’s V7: Video, 3D, and everything we know
Last week’s Midjourney office hours revealed the strategic steps coming soon:
Video Model Release Plans - Midjourney is preparing two distinct video models in collaboration with external partners. The goal is to test community preferences between fast, cost-effective options and slower, high-quality alternatives. The company promises to layer its signature creative elements on top of these models. Initial releases will be experimental to gauge user interest before committing additional resources.
V7 Model: Final Stages of Development - After overcoming some challenges, V7 is back in the final training phase, with a tentative release planned for mid-February. Key improvements include:
Enhanced image quality and prompt comprehension
Stronger multilingual support
A refined character reference system
Potential features for object and logo recognition
New Features Coming Soon - The next 30 to 60 days will bring several updates:
3D Features: Camera movement, reframing in 3D space, and rerendering capabilities
Batch Mode: Large-scale batch processing and a user preference learning system
Post-V7 Enhancements: Faster and slower model variants, along with improved upscalers.
Which one do you want the most?
Read more: Explore Midjourney sref codes discovered by the AI Community
OpenAI’s response to DeepSeek is to launch o3-Mini
Following the game-changing release of DeepSeek, OpenAI has made its move, introducing the new o3-mini reasoning model to ChatGPT and its API, offering enhanced capabilities in science, math, and coding.
Who has access?
Pro users now enjoy unlimited access, while Plus and Team users benefit from triple the previous rate limits. Free users can test o3-mini by selecting the Reason button under the message composer.
Read more: Deepseek R1 is the new big thing in AI. Learn how to use it.
Quick news you may have missed
Luma has introduced Upscale to 4K, enhancing any DreamMachine video up to 4K resolution with all-new detail. You can natively generate in 720p or 1080p, and then upscale to 4K for next-level clarity.
Krea is working to release Krea Chat, powered by DeepSeek(!). This new tool will bring the power of every Krea feature into a chat interface.
Pika has launched Turbo Mode. You can now generate AI videos 3x faster, for 7x fewer credits, without compromising on quality.
Prompt Fight: Golden Stars
Let’s see how the top AI Art generators compare to each other:
Prompt: Beautiful cyborg white woman wearing a black power suit with helmet, made of a slick material with subtle electronic pieces. You can see inside the helmet a golden light like stars. She is standing in a black room illuminated with black golden light, monotone illumination, sci-fi futuristic scenery
Do you have a suggestion for a Prompt Fight? Send me a DM!
Read more: Use 10 prompts for wondrous AI Art
Video AI of the week
The Artist's Journey (@wilfredlee) made the AI short film Waking Fire.
I would contest your overly positive assessment of the Office of Copyright‘s report, it’s a form of censorship justified on an incorrect technical understanding how AI works, in my opinion. Only time will tell of course how material my concern is of course.