🗿 How Brands Use GenAI (Full Breakdown)

GM. Generative AI is improving rapidly. Yet brands are struggling to use it in a way that’s both intellectually protected and fair use.

Most large language models in the market have ambiguous data sources. And as we saw with the NYT-OpenAI lawsuit in late 2023, more content providers and artists will likely start fighting back to be compensated for their loosely volunteered contributions.

Without Further Ado. ☕ *knuckle cracks* ☕ Let’s get into it.

“It’s not you, it’s me”. AI is ready – brands are not.

The technology to produce a Hollywood-worthy film from generated assets and automated tooling is nearly there.

Working with some of our AI pioneering directors at Supernova, we’ve begun to recognize the combination of Midjourney (visual), Eleven Labs (audio), and Luma Labs (editing) as a dream team trio to produce some amazing campaign work.

Brands just aren’t ready to leverage these tools to bring AI-generated advertisements to market.

Mango made recent headlines as the first fashion campaign to use an AI-generated image. But those are basic editorial shots. Not a fully integrated video advertising campaign. And to get that far – Mango developed more than fifteen machine learning platforms (MLE) in-house dating back to 2018.

The majority of brands still need to navigate the regulatory climate before making any big moves.

Here's my breakdown on the latest developments in US copyright law to figure out just how far we are from ubiquitous AI-generated ads.

How the Copyright Act interprets GenAI

The Copyright Act protects human authorship. Not things made from computers. That’s been a precedent for decades. Now that GenAI tools have jumped into the equation, the Copyright Office published guidance stating that “only material that is the product of human creativity” is capable of being protected.

For partial AI-generated creations, only the human-produced parts of the work may qualify for copyright protection IF they contain sufficient creativity. Everything is evaluated on a case-by-case basis.

Here’s one work, entitled “Théâtre D’opéra Spatial”, that was rejected by the Copyright Office.

The artist used more than 600 text prompts in Midjourney to create the above image, then edited it in Photoshop and Gigapixel AI.

The Copyright Office rejected it on the basis that the artist’s text prompts simply “influence[d] what GenAI generated, and that the traditional elements of authorship were determined and executed by the program.”

How should advertisers interpret this? Campaigns created entirely through the use of GenAI tools are not subject to copyright, despite meaningful human involvement or input.

This doesn’t even touch on the subject of fair use, where advertisers could be liable for lawsuits and copyright infringement from the artists and dataset contributors themselves, given most of the software providers are still wrangling how they sourced their language models.

Here are a few ways brands are currently using GenAI to collaborate on marketing and advertising:

1. Brainstorming – use GenAI to come up with a rendering of a concept or visualize a storyboard, then make something original during the design and development stage.

2. Copy – if you use GenAI to draft product descriptions or advertising copy, you are likely not protected. The Copyright Office evaluates whether text is “written entirely by [human authors] without the help of any other source or tool, including any generative AI program”.

3. Generating multiple versions of the same campaign – this is a great way to target unique customer segments or channels by slightly modifying the look, feel, format, or rendering. This is a tricky one from a protection standpoint. If the GenAI tool serves beyond editing, and the artistic elements of the derivative work are a result of GenAI’s contribution, the work is not protectable.

Overall thoughts on brand use of GenAI: AI is clearly useful for marketing & advertising. It makes the production process faster, cheaper, and expands the possibilities of creative expression. For now, brands will have to risk losing protection over their work or infringing on someone else’s.

The capabilities are moving fast, yes. But copyright and IP battles will prevent any revolutionary adoption of text-to-video production in the next 5-10 years in my opinion. Brands simply have their hands tied.

In the next few years, I think the hype will shift from LLMs in the consumer market to enterprise-level intranets designed for large IP holders (similar to Publicis’ Marcel).

s/o Global Legal Post for assistance with the legal advice :)

What We’re Reading

• OpenAI announces SearchGPT, its AI-powered search engine (The Verge)

• Actors go on strike over video games AI threat (BBC)

• 77% Of Employees Report AI Has Increased Workloads And Hampered Productivity, Study Finds (Forbes)

• Introducing Stable Video 4D, The Latest AI Model for Dynamic Multi-Angle Video Generation (Stablity AI)

Reach 200,000+ Tech Consumers and Brands

This is NOFOMO, the newsletter keeping you up-to-date on all things emerging tech — from AI and Blockchain to Virtual Reality.

Advertise with our media conglomerate of thoughtful content creators the digital landscape: X, LinkedIn, Podcasting, Email, etc. We offer exclusive access to a real, engaged community of over 200,000 tech professionals boasting millions of monthly impressions and real results.

Our discerning readers are always looking for the latest tools and products in the industry. Connect with the right people and transform your business.

Request a media kit.

DISCLAIMER: None of this is financial advice. This newsletter is strictly educational and is not investment advice or a solicitation to buy or sell assets or make financial decisions. Please be careful and do your own research.