The future of AI: Cannes Lions 2025 trends you can’t ignore

The top trends you need to know from Cannes Lions 2025 to stay relevant and ready to adapt.

What’s in this report

Chapter 1: The risks of blindly believing bots

Chapter 2: Missing nuance means missing the mark

Chapter 3: Authenticity is the new algorithm

Chapter 4: What marketers really need from AI 

Chapter 5: How to build the smartest strategy in an AI era

This report is your backstage pass to the ideas that defined Cannes Lions 2025. Not the usual round-up of panel soundbites, but a deeper look at what really mattered: the intersection of AI and humanity. The AI wave is rising fast, but as we heard time and time again from marketers, creatives, and strategists alike, it’s the human layer that cuts through.

Foreword

The week’s biggest themes reminded us that relevance in today’s market doesn’t come from scale alone. It comes from staying real, staying curious, and staying connected to the people behind the data. That belief is built into everything we do at GWI. We believe the future of marketing won’t be built on AI alone. It’ll be shaped by emotional intelligence, diversity, and data that reflects the full richness of real people’s lives.

As GWI’s CEO and Founder Tom Smith shared, “AI’s completely turning marketing and advertising upside down. But you can’t make AI work for marketers without high-quality human data.”

If you’re wondering what it takes to build trust, connection, and cultural relevance in this new AI world - start here. This report reflects where the industry is going and how we stay human while we get there.

Let’s dig in.

Key insights

• 1. AI is smart, but it can still get things wrong 

Consumers are flooded with AI-generated content, but that doesn’t mean they trust it. As models keep recycling the same data, hallucinations like made-up quotes and non-existent sources are getting harder to spot. 

• 2. You need real people to get real understanding

Brands are starting to use AI to simulate hard-to-reach audiences, but without real consumer input, these personas can fall flat. When models rely on patchy or biased data, the results can feel off, especially for niche or underrepresented groups. 

• 3. Authentic data makes AI more trustworthy

As AI reshapes how people discover and engage with content, credibility is becoming a major differentiator. That’s why what powers your AI matters more now than ever. With so much machine-made content flooding our feeds, authenticity has become a competitive advantage.

• 4. Marketers need AI that’s fast, transparent, and built on accuracy

In high-pressure environments, speed to insight matters. But it can’t come at the cost of clarity. Marketers want AI tools that simplify complex questions, surface usable answers, and help them move with confidence without second-guessing the source.

Chapter 1: The risks of blindly believing bots

Have you ever read something that sounded smart but turned out to be totally wrong? Welcome to the age of AI, where polish often hides inaccuracy. As AI models scale and synthesize content at speed, their outputs can appear convincing. But as Jason Mander, Chief Insight Officer at GWI, pointed out, “convincing” doesn’t mean “correct.”

At the center of it all is trust and right now one of the biggest threats to that trust is hallucinations - literal flaws in the system where an AI model generates false information like made-up quotes, non-existent sources, fabricated people. 

The trouble is, these hallucinations aren’t happening in a vacuum. As more synthetic content enters the system, the problem only gets worse. Many AI models are now trained on material that was generated by other AIs, creating a kind of echo chamber. Even when the content is human-made, it’s often written by non-experts, or by people making sweeping statements without data to back them up. With each loop, it gets harder to tell what’s real and what’s not. And eventually, those hallucinations stop feeling like glitches and start becoming the norm. 

But, here’s the catch: the output might be automated, but the accountability isn’t. Speaking alongside Jacqueline De Gernier, GWI’s EVP, Commercial at Salon Culture Conversations, Teradata’s CMO, Jacqueline Woods, explained, “It is your responsibility to check those sources. It is your responsibility to understand where this data is coming from so that you can make the determination of whether you trust it or not. Humans have to be a part of this.” 

In other words, no matter how advanced the tools get, someone still needs to unpack where the information came from and if it actually makes sense.

If trust is one side of the AI equation, emotional understanding is the other. And this is where AI still has a long way to go, according to Aarti Bhaskaran, Global Head of Research & Insights at Snap Inc. While large language models can recognize surface-level behavior, they struggle with something far more complex: emotion, intent, and context.

On our Cannes edition of On the dotcast, filmed in Spotify's studio, she points out, “If I were to ask an LLM ‘What do people use platforms for?’, it’s just going to say connecting with friends. But for you to dig a little bit deeper and try to understand the why, it’s actually different. It’s rooted in human emotion and context. You need empathy to understand those differences and I don’t think AI does that. It lacks the EQ that’s needed to sift through data and arrive at an insight.”

That tension took center stage during the conversation where our host and Content Marketing Lead Jenna Kamal unpacked how AI is reshaping trust, truth, and connection. 

But, what does this mean for agencies and brands working at the intersection of AI and audience insight? It starts with building data structures that reflect real people, not just tidy personas. It means layering in things that can’t be scraped or synthesized like empathy, authenticity, and nuance. And most importantly, it means recognizing that human insight is what turns an output into something that resonates.

The end of unquestioned authority

Trust is having a bit of a moment. When anyone can sound credible, how do you spot what’s real? From news feeds and brand messaging to AI-generated everything, the question on everyone’s mind isn’t if AI is smart but if it’s honest.

That concern isn’t just philosophical. It’s practical. As Jacqueline De Gernier pointed out, “30% of consumers and AI adopters say it's the accuracy and the relevancy of the data that's their number one concern.” 

She continued, “We’re talking to a lot of companies where they’re taking our human data and actually building that in so it’s much more representative. I think there are ways you can build solutions in a very responsible way and make sure that the models are being trained on a wide range of diverse data.”

And that matters because how these tools are built shapes how people use them. It’s already reshaping the way brands build connections, platforms serve content, and consumers decide what to believe. Since the release of ChatGPT, consumer concern about AI has skyrocketed by 152%. That shift reflects deeper anxiety: what happens when the tools we trust start sounding smarter than they actually are?

It’s something Cristina Lawrence, EVP of Consumer & Content Experience at Razorfish, has been watching closely. In her view, trust isn’t just a content challenge. It’s a systemic one.

“The model out of the box is not going to get us there. It’s going to tell you what you’re asking for and it’s just going to make assumptions based on the context of your statement,” she explained. “It actually takes a fair amount of work using human insights, survey data, and behavioral data to really get to what it is a person is actually thinking.”

At its core, this reveals we’re still applying old trust to new tech and forgetting that this machine doesn’t always show its sources. The result? Convenience starts to chip away at curiosity.

A recent MIT study makes that risk real. When participants were asked to complete tasks like writing emails or evaluating arguments, those who used tools like ChatGPT were more likely to accept flawed reasoning and less likely to identify errors. The more they relied on AI-generated content, the less cognitively engaged they became and over time, their critical thinking skills declined.  

When critical thinking fades, we stop questioning the very systems shaping our decisions. This moment calls for a new kind of informed decision-making where brands lead responsibly, marketers ground strategy in real insight, and consumers feel the difference.

Chapter 2: Missing nuance means missing the mark

The more AI evolves, the more tempting it gets to fake human connection, especially when real access to niche consumer groups or hard-to-reach personas is scarce. Enter the era of AI-powered personas cobbled together from public data, behavioral guesses, and a trail of digital breadcrumbs. The idea is seductive. If you can’t reach your audience directly, why not recreate them? These digital stand-ins promise quick insight, letting marketers talk to avatars of everyone from Gen Z trendsetters to elusive execs.

Things can get especially tricky when you're trying to understand niche audiences whose lives and motivations often fall outside the bounds of mainstream data collection. When nuance gets left out of the input, it doesn’t show up in the output. That’s where strategies start to miss.

For Jacqueline De Gernier, the solution begins long before the data is fed into the model.

“We absolutely need to lean in,” she shared. “It starts with ensuring that we’re building diverse teams.”

As Lisa Smith, Partner, Marketing Emerging Technologies at Deloitte Digital UK, reminded us, it’s about approaching it with integrity. She shared, “I think the best thing you can do is be honest, so that people are courageous enough to interrogate the data and to know that maybe they are deploying something that’s imperfect.”

Many of the conversations at Cannes Lions  this year circled the same issue of representation and the same question kept resurfacing: who gets to be seen in AI? It’s not just about who shows up in the data. It’s about who gets left behind in the decisions that shape strategies, campaigns, and products.

For Aarti Bhaskaran, this isn’t just theory. It’s a lived experience. “I asked one of the AI tools [to generate an image of me] and my avatar was an Indian man because the training data is very much gender skewed. So, the moment you say I’m a professional of a certain age, it skews towards male.”

It’s a small moment with big implications. The problem is that a lot of the data that trains AI already carries bias and unless someone steps in, that bias just keeps getting baked in. 

She continued, “I’m wondering how many more instances has this been thrown out and then people are having conversations with these avatars and making decisions.”

The fine print of inclusion

Bias is an enormous problem in the world of AI. What looks neutral on the surface can hide a world of assumptions underneath. And too often, those assumptions leave entire communities - especially minorities or marginalized communities - out of the frame.

Jacqueline De Gernier explained “There’s a really concerningly small number of people that are responsible for defining the data that the AI models are trained on. There are 6,000 languages globally and yet most of the data that the models are trained on, their sources are in English. That immediately sends off a warning signal.”

In AI-driven marketing, absence isn’t just neutral, it’s a design flaw. And it’s one we can’t afford to ignore. If the models can’t see us, the strategies built on them won’t either, leading to campaigns that miss the mark and messaging that misfires.

We saw that lens in action at Cannes Lions at The Insight Table, GWI’s panel and lunch that brought together women in leadership for bold, honest conversation. It was a candid look at burnout, bias, and the emotional math women in leadership still have to do. It was a reminder that data can’t tell the full story unless the people shaping it are heard first. 

Inclusive AI doesn’t happen by accident. The fix starts upstream with better data, broader perspectives, and inclusive teams shaping the tools from the ground up. It’s not just what we build, it’s who gets to build it. Because when the foundation is built for everyone, the results speak to everyone. That’s how we move from oversight to insight.

Chapter 3: Authenticity is the new algorithm

As AI content floods every feed, it’s important to remember that the real currency is trust. Why? It’s how you stand out in the flood of machine-made messages. Authenticity is a strategic edge. 

But it isn’t something you can just prompt into existence. As we’ve mentioned, it has to be built into the process from the data you use to the stories you tell. That mindset is starting to reshape how marketers think about brand visibility.

“Search is changing as consumer behavior is changing,” Bhaskaran explained. “The search is moving from knowledge to outcome. So, you’re not buying by brand, name, or top keywords. You have to think of outcomes and make sure you present them.”

Enter GEO - generative engine optimization. If traditional SEO was about ranking on Google, GEO is about ranking in AI. In this new reality, what matters most now is how you show up in the wild.

As Jenna Kamal puts it, “SEO is your own house. You design it. You decorate the rooms. You put your name on the letterbox. If someone visits, they get the exact experience you’ve curated. Your website is that house. You write the content. You optimize the structure. You decide the keywords, the headlines, and the internal links. You’re in control of what people see - and how it performs.”

She continues, “GEO - also known as LLM SEO -  is your reputation in the neighborhood. It’s what people say about you when you’re not in the room. It’s the flyers someone else put through the door. It’s the blog post a journalist wrote mentioning your brand. It’s the Reddit thread where users debate your product vs. a competitor. It’s the TikTok review that went viral without you lifting a finger.”

Jason Mander sees this shift playing out in real time.

“We know people crave authentic recommendations when making decisions, whether personally or professionally. We want validation and reassurance from our tribes and our peers, people we know and can trust. That’s one of the reasons the GEO space is so interesting because it’s where performance is directly impacted by what people are actually saying about your brand, about reviews, and about how human your data feels.”

As the role of AI grows, so does the need to stay grounded in what makes a brand feel real.

Jason Mander continues, “In the future it seems clear that authenticity is going to matter more and more because what powers your tech and AI is going to be every bit as important as what that tech can do.”

Marketers know this intuitively. The biggest benefits they report from using AI tools aren’t flashy features. They’re time efficiency, productivity, and cost savings. But none of that matters if the trust breaks. That’s why what powers your AI (and how human it feels) is becoming a brand differentiator.

If trust is the new currency, then authenticity is how you earn it. Brands can’t afford to be surface-level. The inputs matter, the sources matter, and the more human your footprint, the more lasting your impact.

The real you is the real advantage

Even though AI can replicate style, format, and even tone, it can’t replicate the magic of being fully human. Everyone’s got access to the same tools, but no one else has your lived experience, your voice, or your lens. Showing up as yourself is one thing AI can’t replicate.

The next big marketing advantage is personal. You just need to show up as yourself - clearly, consistently, and with conviction. People aren’t just choosing products. Consumer decisions have become a form of self-expression, shaped by identity, belief, and belonging. They gravitate toward brands that reflect their worldview, mirror their values, and feel human in the process. That explains why 46% say they want brands to be authentic. 

But when many marketers are using the same tools, there’s a growing risk of convergence. 

As Anna Grodecka-Grad, SVP, Global Client Services at StackAdapt, warned, “If different companies, agencies, and tech are using the same AI tools, we may actually get the same answers to our prompts. And then that becomes personalization that’s not personalized and not relevant. Sameness is a big risk and that's where I think the AI and human combination is so important.”

In the age of AI, humanity is your winning strategy. When everything feels artificial, honesty becomes magnetic. What sticks are the moments that feel real, grounded, and intentional. That’s not always easy to pull off, but it’s what builds trust over time.

⭐ Cannes Lions spotlight: Unfiltered and unscripted

At Cannes Lions 2025, GWI came with insights. But we also came with real moments, unscripted interactions, and fun bloopers. Behind every polished stat and big idea is a team of humans just doing their best. We laughed, we flubbed our lines, we asked for one more take. Honestly, those moments said just as much as the panels did.

As a brand, showing up as your real, unfiltered self might be the boldest strategy of all. Why? Because it’s the fastest way to spark genuine connection and the kind of engagement you can’t fake.

To prove the power of staying human, we hit the pavement with a mic and a little pop quiz action with Data on the Streets. It’s our way of bringing insights to life by asking real people real questions, grounded in GWI data.

It’s all about bridging the gap between human truth and machine-made content. Whether it’s on a main stage, on the sidewalk, or online, the brands making the biggest impact are the ones leaning into what makes them real.

Chapter 4: What marketers really need from AI

AI is everywhere, but the tools that stand out aren’t just faster - they’re smarter. But, marketers don’t need more AI for the sake of it. Our data shows that the top concerns holding teams back from adopting AI include privacy, lack of trust, and lack of expertise. They need AI backed by insight they can trust and that actually helps them make sharper calls by filling in their skill gaps. Speed means nothing if the answer isn’t useful and automation doesn’t help if you’re still left guessing. That’s exactly where GWI Spark comes in.

GWI Spark is an AI-powered insights assistant trained on real survey data from nearly a million consumers across 50+ global markets. That means every answer is rooted in verified behaviors, opinions, and preferences. Whether you’re planning a campaign, shaping a pitch, or just trying to understand a new audience, you’ll get clear, credible insights in seconds. 

As Cristina Lawrence reminded us, the real power of AI lies in what fuels it.

“What makes AI incredibly useful is if you have human-centric insights and data. It’s a terrific tool for extending the life of that data. You’re able to actually interact with it real-time. Instead of combing through pages and pages of insights, you could actually ask the question and get what you need back.”

When you combine credible data with conversational AI, you unlock a new way to make insight more accessible, more actionable, and more enduring across teams.

GWI Spark isn’t just for analysts. It’s for anyone who needs to move fast with confidence. For marketers, it’s a shortcut to proof points that turn ideas into action. For strategists, it’s a way to validate assumptions and explore new angles without missing a beat. And for product teams, it’s a way to understand what people need before you build. 

Here’s how it works: You start with a question - anything from “What motivates Gen Z to switch brands?” to “How do gamers feel about in-game ads?” GWI Spark reads that question, understands the topic, identifies the audience, and determines what kind of insight you’re actually looking for. Then it searches millions of real survey responses from our global data set, filtering for relevance, strength, and statistical significance.

Every insight is scored across three criteria: compatibility with your question, index strength (how strongly a behavior or opinion stands out), and prevalence across the audience.

Finally, it ranks the most valuable insights and gives you a curated breakdown of your audience, ready to plug into your pitch, campaign, or strategy deck.

Traditional AI might give you speed, but GWI Spark gives you substance. So when it’s time to back up a bold idea or defend a decision, you’re standing on solid ground.

Every idea starts with a spark

What drives great ideas? For GWI, it always starts with insight. At Cannes Lions 2025, we brought that idea to life at the GWI Stadium to show how fast, trusted insights can ignite your next big move. It was part demo, part destination, and all about giving attendees a hands-on look at what GWI Spark can really do.

At the heart of the space was our interactive GWI Spark setup where visitors could ask real questions, explore live insights, and see how our AI tool turns curiosity into clarity in seconds. We turned data discovery into a race to see who could uncover the sharpest insight the fastest.

With just a question and a few clicks, they explored real consumer data, zeroed in on trends, and uncovered audience truths in record time. It was proof that when speed meets substance, strategy gets a serious upgrade - and they loved it.

From brand strategists to creatives and CMOs, everyone who stepped into the GWI Stadium walked away with one thing: a real sense of what insight at speed really looks like. It was live, hands-on, and showed just how quickly GWI Spark can surface the stories behind the stats.

And that’s important in an industry where timing and relevance are everything. Fast data isn’t new. But fast data you can trust? That’s the game-changer.

Chapter 5: How to build the smartest strategy in an AI era

54% of workers now use AI and another 21% are interested, showing just how quickly these tools are reshaping the way we work. It’s tempting to move fast, to plug in prompts, to chase whatever’s trending. But real strategy starts with understanding real people and that comes from trustworthy data. That’s the line between a flash of attention and long-term impact.

As Tom Smith says, “Insights are a superpower for any business. They spark ideas, fuel innovation, and unlock new opportunities. They’re the foundation of great strategy and the key to building a lasting competitive edge.” 

When you know what people care about, what they value, how they behave, and where they show up, everything gets easier. Your creative lands harder. Your pitch hits sharper. Your product roadmap gets clearer.

So, how do you actually build smarter strategies in an AI-powered world? 

1. Question the source: Not all AI is created equal. If you don’t know where the data comes from, you can’t trust the output. Insight without origin isn’t insight,  it’s a guess. Build your strategy on verified, high-quality data that reflects real people, not assumptions.
   
   
2. Don’t just prompt - probe: AI can surface patterns, but it takes human curiosity to turn those patterns into meaning. If your inputs lack diversity, your output will too and it’ll show. The most resonant strategies start with real representation: voices that reflect different cultures, backgrounds, and lived experiences. 
   
   
3. Treat insight as a team sport: Insight shouldn’t live in isolation. When everyone from creative to strategy to product can access and explore consumer data, ideas evolve faster and strategies get sharper. Share the spark.
   
   
4. Make speed work for you, not against you: The pressure to move fast is real, but speed means nothing without clarity. Don’t trade velocity for vagueness. A great strategy needs both momentum and meaning. Tools like GWI Spark give you credible insights in seconds, without sacrificing depth. 
   
   
5. Stay human: Impactful strategies don’t just check the data box, they reflect the nuance, empathy, and cultural awareness only real people bring. Real connection still comes from real people and that’s exactly what the future of marketing needs.

As Jason Mander said, “Human voices are the anchor that give the depth, the complexity, the nuance. That’s what gives authenticity to the data and is going to be what builds trust. 

Every strategy needs human understanding and that comes from real people, real voices, and real data.

It’s not just about what your tools can do. It’s about what you do with them. The smartest strategies in the AI era will come from marketers who stay curious and stay grounded in reliable insights. When your tech serves truth, not just trends, you build something that lasts.

Tom Smith put it best. “AI is incredible but it’s very clear that it’s lacking the human element. Most people are using out-of-the-box LLMs or synthetic audiences. These are prone to hallucinations, bias, and the insights you get are inaccurate. So, you need the best AI and the best data. That’s what GWI is committed to bringing together going forward.”