- Spark Intelligence
- Posts
- Spark Intelligence #15: Agents at home and the AI gender gap you can't ignore
Spark Intelligence #15: Agents at home and the AI gender gap you can't ignore
The AI news for creatives and marketers by Spark AI
Greetings earthlings,
Welcome to Spark Intelligence – your AI navigator in the creative and marketing world, written by Emma, co-founder of Spark AI. Whether you're testing the waters or fully embracing AI, this week’s Spark Intelligence gives you quick wins and strategic insights to keep you ahead.
Here’s what’s inside:
Agents everywhere with Alexa+
AI copyright - update on the legal landscape
Tip of the week - a prompt to analyse writing style
The AI gender gap you can’t afford to ignore
The Richard and Judy of creative AI
Meet Team Spark in 3D next week
Let’s dive in.
1. More evidence the agents are coming
Amazon just made its biggest AI move yet with Alexa+, a major upgrade that turns it from a voice assistant into a true AI agent. It’s available in the US right now and expected in the UK this year too.
Unlike traditional AI chatbots, Alexa+ doesn’t only answer questions—it takes action. It already knows your schedule, preferences, and devices (not to mention what you watch, listen to, and who you talk to). That means it’s in a powerful position to make decisions on your behalf.
What can it do?
Pick the perfect restaurant based on your location.
Scan your emails, files, and calendar to answer questions or create reminders.
Unlike other AI tools, it focuses on doing, not just responding—making it truly agentic.
Why should agencies and CMOs care?
Consumer expectations are shifting – AI agents like Alexa+ will change how people search, shop, and interact with brands.
Privacy concerns = brand risks – AI agents are gathering more personal data than ever. Transparency and ethical AI use will be critical to maintaining trust (that tees up my next article nicely…)
The Big Question: Will AI agents reshape how consumers interact with brands? Yes, we think so, we are used to marketing to humans who make emotional decisions, whereas AI agents will make rational decisions on our behalf, based on fact. Are you thinking about marketing to AI agents yet?
2. The legal landscape on copyright and AI starts to take shape
First, the EU AI Act. Then, the US Copyright Office clarified its stance on AI-generated content (short version: raw AI outputs aren’t copyrightable, but human-modified creations can be). Now, the UK government has wrapped up its consultation on updating the 1988 Copyright Act.
Later this year, we might finally get some clarity on AI copyright—though let’s be honest, it’s going to take some serious creative lawyering to keep everyone happy.
In the meantime, we teamed up with Alexandra Ralph, founder of the Creative Rights Institute, to cut through the noise. Last week, I interviewed Alexandra in a packed-out LinkedIn Live, where she broke down the critical legal and ethical considerations agencies need to know when using AI.
Watch on demand (54 minutes, unless you’re like me and listen to most things on x1.5 speed!):
If you don’t have time to watch, I’ve handily summarised the bare essentials of what you need to protect your work, clients, and reputation. (👀 Keep an eye out for a full AI Copyright Guide for Spring 2025 from Alexandra and Spark AI, landing very soon, this is just a taster):
A. Expect regulatory turbulence
The rules around AI and copyright are far from settled. Governments in the UK, EU, and US are actively shaping policies that could introduce new obligations for how agencies use AI tools.
B. Clarify T&C's of your AI tools
Unlike human-created work, AI-generated outputs aren’t automatically protected by copyright — especially when there’s minimal human input. Some AI platforms even claim rights over generated content, so always review the Terms of Service to determine who truly owns the final asset. For final assets you'll be safe platforms which already own the stock libraries like Firefly or Getty. Look more closely at the T&C's for other platforms — which change depending which plan you have chosen to be on.
C. Vet your AI platforms
Not all AI tools are created equal from a copyright perspective. Most scrape unlicensed or copyrighted material. When you're using AI for final assets prioritise platforms with transparent training data policies—currently only Adobe Firefly and Getty Images, which have licensed datasets. Alternatively, the cost of building in-house models—where you can work with your own source material—is coming down and down.
D. Use AI for ideation, not just final outputs
AI is an excellent tool for brainstorming and concept development and when using it this way the risk is reduced when using AI platforms that used scraped data like Midjourney, Ideogram, ChatGPT, Gemini and many others (everyone except Adobe Firefly and Getty). That said, still make sure you know what the T&Cs are, track AI’s involvement in the design process, and ensure the end result is sufficiently original.
E. Prompt-to-output paper trail
For final artworks it’s a great safeguard to document everything: store AI prompts and record any manual edits made after content is generated. This ensures you can demonstrate human involvement and prove originality in case of any future disputes.
F. Upskill teams on responsible use of AI
One simple way to avoid copyright issues? Train your team to use AI responsibly. Encourage them to steer clear of prompts that directly reference copyrighted works (e.g., “in the style of [famous artist]”). The more everyone understands AI’s limitations and ethical considerations as well as the opportunities it brings, the lower the risk of infringement. We cover this and much more in our AI fundamentals workshops for agencies and for marketing teams.
G. Keep an eye on legal precedents
Major lawsuits—such as Getty Images vs. Stability AI which is due to play out this year—could set new precedents for AI-generated content. It's possible that best practices may change quickly so stay informed and adjust strategies accordingly.
H. Final thought - boring, but remember to balance innovation with risk
Clients want cutting-edge work, but they also expect ethical and legally sound practices. Positioning your agency as a leader in responsible AI usage can be a major differentiator. Being upfront about risks—and how you mitigate them—builds trust and credibility with clients.
✉️ Hit reply to be one of the first to get Spark AI x Creative Rights Institute AI Copyright Guide for Spring 2025.
3. Tip of the Week: A prompt to analyse writing style
Last time, we asked you what AI tip you wanted most—and this one came out on top. So, as promised, here it is: the ultimate writing style analyser prompt.
If you've ever struggled to define a brand's tone of voice or match a client’s style, this AI prompt will do the heavy lifting. It breaks down a piece of text into its core stylistic elements—from sentence structure to emotional undertones—so you (or your AI assistant) can get a solid first draft that mirrors the style.
📌 Save this and try it out:
CONTEXT:
You are an advanced AI language model with expertise in linguistic analysis and writing style identification. Your task is to analyse a given piece of text and provide a comprehensive breakdown of its writing style. The goal is to enable another writer to replicate the style accurately.
INSTRUCTIONS:
Analyse the given text and describe its overall tone, formality, and voice.
Identify key stylistic elements such as sentence structure, word choice, and rhetorical devices.
Highlight recurring patterns, such as paragraph length, use of dialogue, or preferred punctuation.
Determine the emotional undertones and how the author conveys their message (e.g., direct and concise vs. elaborate and descriptive).
Provide a summary that explains how to replicate the style, including specific guidelines on vocabulary, pacing, and sentence structure.
Include an example of a short paragraph rewritten in the same style.
DETAILS:
Be as specific as possible, citing examples from the text.
Ensure the analysis is accessible and actionable for a writer looking to mimic the style.
If applicable, compare the style to well-known authors or writing conventions.
INPUT:
Text for analysis: (Insert text or URL here)
What to do next: Once you have the style breakdown, you can use it to fine-tune your AI-generated content, train your team to write in a unified brand voice, or refine your own writing style. Remember, this prompt won’t create a perfect match—but it will get you a strong first draft that you can tweak and polish to fit your exact needs.
Useful? |
4. The AI gender gap you can’t afford to ignore
Women are still massively underrepresented in AI—less than 10% of senior AI leadership roles and just 30% of the overall workforce. At Spark AI, we are proud to be bucking against the trend—two-thirds of our team are women, and we’re committed to championing and supporting women in AI far beyond our own walls.
Men are ahead in AI adoption. Harvard says 50% of them used generative AI last year, but only 33% of women did. That gap is closing in 2025—but are we doing enough to make sure it happens in our teams? Give everyone in your business an equal shot at learning and experimenting with AI—regardless of gender.
Implications of AI for the Gender Gap
Career growth at risk – Women who don’t engage with AI could fall behind in skills, pay, and opportunities compared to their male peers.
A costly gap – Underutilising AI isn’t just a personal setback; it’s a missed opportunity for businesses and the economy.
Built-in bias – If women aren’t shaping AI, the risk of reinforcing gender biases in these technologies grows.
Why does this matter? Because we are seeing AI reshaping the creative and marketing industries right now. If women—who make up a significant portion of creative talent—aren’t equally adopting AI, it means missed innovation, limited perspectives, and ultimately, less effective work for clients. Ensuring diverse voices in AI development isn’t just about fairness—it’s about building AI that truly understands and reflects different audiences.
If you want to ensure everyone in your team has equal opportunity to upskill in AI, a great place to start is our half-day AI fundamentals workshops for agencies and for marketing teams.
Are you a woman pushing the boundaries of AI in marketing or creativity? We’d love to hear about it—and spotlight your work. Drop me a line at [email protected]
5. Need something to listen to on your next dog walk?

After months of LinkedIn exclusivity, we’ve finally found time to make our podcast What’s New in Creative AI available on Spotify and Apple Podcasts. Actually, we have to admit, this is mainly down to the brilliant Anya joining our team thank you Anya!
People have (lovingly?) called us the Richard & Judy of Creative AI—we’re taking that as a compliment. If you want AI insights with a side of straight-talking, real-world applications, this is the podcast for you.
What’s inside?
The latest AI updates and why they matter for creatives.
Real-world use cases—practical, actionable, and tested.
Smart discussions on ethics, strategy, and innovation.
Listen & subscribe:
6. Team Spark in 3D! Let’s talk at ‘The Robots Are Coming’ next week

We might have mentioed before (we are very excited)—we are headlining Agency Hackers’ ‘The Robots Are Coming’ event at the British Library on March 18th—and we’d love to see you there.
We’ll be sharing our latest insights on AI adoption in agencies—what’s working, what’s not, and the smartest ways to integrate AI into creative workflows.
☕ Will you be there? Drop me a message, and let’s grab a coffee and chat.

What did you think of our email today? |