It’s been a breathtaking couple of years since the rise of AI, and I hope you survived the onslaught. In other words, if you’re a creative, you’ve fought off any specific threats to your livelihood from generative AI, and are still managing to earn money.
Even if that’s the case, though, it doesn’t mean you can rest easy. Because AI continues to evolve at breakneck pace, and there’ll be little time in 2025 to catch your breath, before everything changes again.
I’ve written extensively on AI and creativity over the last two years, and interviewed a lot of people involved in the field. So while it’s difficult to make specific predictions, I do at least have a broad sense of where things are going. Read on, as I share some thoughts about what to expect in 2025. For a look back over the last 12 months, take a look at the biggest AI controversies of 2024.
Enter the Controls Era
We’ll start with what could be a game-changer for creatives. AI-generated art has so far relied heavily on text prompts—a mechanism that many professionals have rightly criticised as reductive.
However, 2025 promises to usher in a new “Controls Era” in AI, where people can steer the outputs of generative tools with more precision.
The idea here is that instead of simply describing what you want, you’ll be able to directly adjust elements of AI-generated content, whether that’s tweaking the lighting in a generated film scene or refining brushstrokes in an AI-assisted illustration.
Is that a good thing or a bad thing? Well, it largely depends on whether you like using generative AI in your work, or you’re worried about non-creatives using it and putting you out of a job.
Adobe‘s Alexandru Costin feels you should veer towards the former. As he puts it in this interview, “Think of these ‘controls’ as the digital equivalent of the paintbrush in Photoshop. This tool once seemed threatening to traditional art forms but ultimately sparked an explosion of creativity, craftsmanship, and innovation. Photoshop didn’t replace physical paintbrushes; it redefined what was possible, opening new avenues for precision, taste and expression.”
In other words, the main threat of generative AI has been how easy it is for non-creatives to use them. But by adding more layers of skill in this new “control era”, the AI pendulum will swing back to professional artists.
To put it in blunt terms, clients will hire us because we can “use AI”, just like we once could “use Photoshop“. In neither case was this actually what we were doing—it was about our creativity, not our software expertise—but as long as the money went in the bank, we were never going to argue about it too much.
Specialised creative tools
Here’s something else that could help reposition AI as a creative skill, rather than a creative threat: the proliferation of AI tools tailored to specific disciplines.
Imagine a tool for illustrators, for example, where AI assists with intricate detailing while respecting your own personal artistic style. Similarly, I’m expecting new AI tools to help game designers build more immersive worlds with realistic NPCs (non-player characters).
Then there’s video. Platforms like Sora and Meta’s Make-A-Video are already pushing the boundaries of generative AI that moves. And while we’re some way from useable content from these platforms, I’m expecting to see an AI platform for film-makers that helps to simplify pre-production processes.
Scene layout tools that simulate lighting and camera angles, for instance, could dramatically reduce the time you’d need to spend on setting up a scene. Then there’s animation, where maybe AI tools could turn storyboards into animated sequences within hours. (For more on this, read our article The future of AI in filmmaking – from storyboard to screen.)
None of these tools exist yet, I should add. But from what I hear, they’re all in the works. Will they get released in 2025? Maybe, maybe not. But as a broader point, AI tools will certainly get better and better at handling tedious tasks, freeing creatives to focus on strategic and emotional storytelling.
Clearly, not everyone will get on board with this. But I do believe that, whether it be for good or bad, more and more creatives will come to see AI as a collaborator, rather than a replacement, in 2025.
Court cases and legislation
Before you get too excited, though, there’s also a possibility that AI could come crashing down in 2025. Because despite the huge power of the tech industry, they’re still not above the law, as demonstrated by the dramatic arrest of Telegram CEO Pavel Durov in France this summer.
While AI is amazing, it’s worth remembering that it’s trained on real-world content, owned by people who mostly never gave permission. As a result, several major lawsuits are underway in both the US and UK, with companies like OpenAI, Microsoft and Anthropic facing numerous copyright infringement claims. Should a single judge come down on the side of the complainants in any of these cases, it could quickly establish a legal principle that makes AI software very difficult to train.
For governments, meanwhile, deepfake regulation is becoming a priority, with rising concerns about the use of generative AI for purposes such as pornography, fraud and political disinformation. Tennessee has already passed the ELVIS (Ensuring Likeness Voice and Image Security) Act to protect performers’ likenesses. And the EU, Japan, Israel and Singapore have all introduced legislation affecting AI training using copyrighted material.
It’s unlikely, though, that any of this will stop AI in its tracks entirely. For a start, court cases take a lot of time to come to fruition, and turning legislation into effective regulation and enforcement can also take aeons. So 2025 might be largely a year for kicking the can down the road.
Even if that’s not the case, I don’t think it’ll be devastating for the AI firms. For example, I’d expect to see a requirement to provide opt-out mechanisms for copyright holders and improve AI’s ability to handle abstract concepts without drawing on specific copyrighted material. But while these limitations would frustrate some tech leaders, they’d definitely be doable.
Learning to live with AI
Many of the people reading this article will want them to go further. Some would love AI to be banned outright. But ultimately, we have to be realistic. Trillions have been poured into this thing, tech leaders are swarming all over the next White House, and quite simply AI is not going away. Which means that, just like Covid, we’ll have to learn to live with it.
Whether that means you co-opt AI, collaborate with it, or simply ignore it, is of course up to you. After all, plenty of artists still slap paint onto physical paper, plenty of photographers shoot on film, and I even know writers who still use mechanical typewriters.
At the end of the day, clients don’t pay you for the tech you use; otherwise everyone with a pricey SLR would be a pro photographer. So always remember, they’re paying you for your creative mind… and no tech billionaire, and no clever technology, can ever take that away.