Cannes Lions 2026: It’s Time to Move from Brand Safety to Audience Safety

Cannes Lions is where our industry celebrates creativity, effectiveness and the commercial power of ideas. It's also, increasingly, where we corner each other on the Croisette to ask harder questions about the systems underneath all of it.

This year, the question I keep coming back to isn't whether media environments are brand-safe. It's whether they're audience-safe, and whether the standards we apply to platforms have caught up with what audiences already expect from us.

They haven't.

For years, brand safety has been a proxy for risk management. Keep your ad away from the beheading video and the conspiracy theory, tick the box, move on. That framing was always self-serving. It centres the brand rather than the people the advertising actually reaches. And in a media environment where children and teenagers are routinely sitting in spaces built for adults, it doesn't hold up anymore.

The shift we need to make is from brand safety to audience safety. Audience safety asks a different question. Not “is this environment clean enough for my brand?” but “who is actually in here, and what's happening to them?” It explores how the platform is built, what gets amplified, what gets through moderation, not just what your ad happens to sit next to.

And it counts everyone in the room, including the people you never meant to reach. The people you “never meant to reach” are not marginal; often, they are sitting in the same feeds, apps and devices as the audience you paid to reach. That is where the old brand-safety frame starts to break down most clearly.

This is where the conversation usually gets vague, so let me be specific. A media buyer told me recently, half-laughing, that her seven-year-old watches more YouTube on her phone than she does. That's the room we're actually planning into. Children are sitting inside adult media environments far more than advertisers like to admit. Parents aren't always there. Devices get shared. Younger kids watch what's made for teenagers; teenagers live on platforms built for adults. The tidy idea that you can separate a “kids” environment from a “family” environment from a “general audience” environment is, in practice, a fiction.

This matters for a simple reason: if you do not know who is actually in your audience, you cannot make responsible decisions about where you spend. Family-facing brands often assume that reaching the category buyer (usually a parent) means they're in safe territory. But that parent and their kids tend to share the same screen, the same apps, the same feed. The household media environment is messier and more overlapping than any media plan I've seen wants to admit.

And this isn't only on family brands. If you spend at scale on a major platform, some share of your audience will be children or teenagers, whether you targeted them or not. That makes audience safety a systems question, not a category question. It is also why trust and safety can no longer sit in a specialist corner of the business.

Trust and safety used to be a specialist's problem, something the policy team or compliance worried about, while the rest of marketing got on with the real work. That is no longer the case; the shift isn't always happening for the right reasons, and it's slower than it should be, but the centre of gravity has moved.

Today it sits right where media investment, brand reputation and regulatory exposure meet. The Digital Services Act, the Online Safety Act, and the Children's Code all materially impact digital advertising. They're live obligations that shape where you can spend, what you can say and who you can reach. Brands that treat this as paperwork will end up on the wrong side of the next news cycle, whereas the ones that build real capability here will pull ahead, and stay ahead.

If audience safety is the frame, then advertisers need to be asking different questions. The new questions need to be focused on age assurance, moderation, escalation and accountability. Before committing spend, ask:

• On age assurance and exposure: how does the platform assess and manage underage users? What safeguards exist where children or teenagers are likely to be present? How does age-appropriate design affect content distribution and ad placement? Do not accept reassurances. Ask for the methodology.
• On content moderation: what systems exist to identify and remove harmful content? How effective are they in practice, not in theory? What does failure actually look like, and how often does it happen? The answers to these questions are rarely published voluntarily.
• On ad placement and transparency: what control do advertisers have over where their ads appear? How is adjacency defined, monitored and reported? “We use industry-standard brand safety tools” is not an answer.
• On escalation: what happens when something goes wrong? How quickly are issues resolved? Who is accountable? If a platform cannot answer this clearly, that is itself an answer.

I sat in a room recently with a brand team going through this exercise properly for the first time, and the silence after they asked the adjacency question was the most honest thing in the meeting. Nobody on the platform side had a clean answer. Nobody on the brand side knew what to do with that. It's the kind of moment that makes you realise how much of this conversation is still happening in the abstract, and how rarely the right questions actually get put in front of the right people.

And adjacency is the comparatively simple version of the problem. If platforms struggle to explain where an ad appeared, what sat around it and who was accountable when something went wrong, advertisers should be much more sceptical about their ability to govern media environments being reshaped by synthetic content, automated optimisation and generative systems operating at scale.

Cannes 2026 will be drowning in AI. Generative creative, synthetic media, AI-optimised targeting. Most of it gets sold as progress, and some of it genuinely is. But there's a set of trust and safety questions underneath all this that the industry keeps dancing around, and advertisers need to start pushing on them.

When a platform hosts AI-generated content at scale, moderation becomes a fundamentally different problem. The volume is higher, it moves faster, and there's no human author to escalate to. The old rules weren't built for that. Advertisers should be asking, plainly, whether the platforms they buy on have actually rebuilt their moderation and enforcement for synthetic media, or whether they're just running last decade's playbook with a new label on it.

So: how are platforms labelling AI-generated content? What provenance standards apply? What's the takedown process when synthetic media causes harm, and how does that process hold up when the content involves children or young people?

These aren't hypothetical questions. The harms are already documented. Set our industry's silence on this next to its enthusiasm for AI-generated creative, and the contrast doesn't reflect well on any of us.

By Cannes 2027, here's what I'd like to be looking at.

First, advertisers asking the hard questions before they commit spend, not after the crisis hits. The reactive playbook (pull the ads, issue the statement, wait for the story to move on) is expensive and it doesn't work. Doing the work upfront is more ethical and, frankly, better business.

Second, family-facing brands leading instead of following. If you market to households with children, the duty of scrutiny on you is higher. Not just because the regulator says so, though increasingly it does. Because some of the most vulnerable people in any media environment are sitting inside your audience, and they can't speak up for themselves.

Third, the industry treating audience safety as something it actually does, not something it buys in. The brands that build real capability here, the ones that understand how platforms work, what enforcement actually looks like, and what their audiences are dealing with day-to-day, will make better decisions. And better decisions, over time, are what trust is built out of.

This is part of why I've been spending time with the Online Responsibility Network. ORN is doing something the industry has needed for a long time: a place where organisations can actually compare notes on what trust and safety looks like in practice, get independent assurance on how they're doing it, and stop pretending each of them is solving these problems alone. I've been helping test the new portal as it gets off the ground, and the quality of the conversations on it is the most useful signal I've seen this year that the centre of gravity really is moving. If you're a brand or platform serious about audience safety rather than the PR version of it, it's worth a look (onlineresponsibilitynetwork.com).

The platforms aren't going to solve this on their own. They've shown us that. The question is whether advertisers will finally use the leverage they've always had.

They're the ones writing the cheques. Time to start asking what they're paying for.

Sev Marcel is COO at We Are Family, a global agency specialising in kids, teens, and families. With over 20 years in the industry, she brings together brand strategy expertise, research rigour, and the hard-won instincts of a working mum to get under the skin of what genuinely drives young audiences and earns a lasting place in family life. Known for her sharp, direct style, Sev cuts through the noise to get to what actually matters, whether that's in the strategy, the research, or the room.