When Safety Became Alignment

How a sensible framework for avoiding harm became a subjective filter for cultural comfort, and what it is costing advertisers.

There was a time when brand safety meant something specific. It meant not appearing next to content that was genuinely harmful. Hate speech. Incitement. Illegal activity. Graphic violence. The principle was simple and hard to argue with. No serious advertiser wanted their logo beside something that could cause real damage to real people.

That was a sensible framework. It served the industry well for years. And then, gradually, it stopped being the framework anyone was actually using.

The shift did not happen overnight. It happened through accumulation. A few high-profile incidents, most of them on YouTube between 2017 and 2019, created a climate of anxiety. Brands discovered that programmatic buying could place their ads in places they had never intended. The response was understandable. Exclusion lists were created. Keyword blocklists expanded. Safety verification became a requirement on every media plan.

So far, so reasonable.

But something else happened alongside the operational fixes. The definition of what counted as unsafe began to stretch. It moved from content that was harmful to content that was controversial. Then from controversial to uncomfortable. Then from uncomfortable to simply unfamiliar.

Risk frameworks that were designed to prevent genuine reputational damage started being used to manage cultural association. The question stopped being “is this content dangerous” and became “does this content reflect our brand values.” Those are two very different questions, and the industry has been answering the second one while pretending it is still answering the first.

Brand safety became brand alignment. And very few people noticed the swap.

This matters commercially, not just philosophically. When you exclude media environments on the basis of cultural alignment rather than genuine risk, you are removing audiences from your addressable market. Not because those audiences do not buy your products. Not because the environment fails to deliver attention or recall. But because someone in a meeting decided the association felt wrong.

The compounding effect of this is significant. Each exclusion narrows reach. Each narrowing increases the cost of reaching the remaining audience. Each cost increase pressures performance metrics. And each underperformance reinforces the argument that the excluded environments were not worth including in the first place. It is a self-fulfilling loop, and it runs on assumption rather than evidence.

Meanwhile, the audiences being excluded have not disappeared. They still buy cars. They still take out insurance. They still do the weekly shop. They still book holidays. They are commercially active, reachable and, in many cases, less expensive to reach than the audiences everyone is competing for in the environments that remain on the approved list.

There is an additional layer that makes this more uncomfortable. Media consumption is increasingly treated as a proxy for belief. If someone watches a particular channel, reads a particular newspaper or follows a particular account, assumptions are made about who they are and what they think. Those assumptions are then used to determine whether they are brand-safe audiences.

This is a genuinely risky shortcut. People’s media habits are not a reliable indicator of their purchasing behaviour, their values or their character. The evidence on this is fairly clear. Category buyers are distributed across the entire media landscape, not concentrated in the environments that make agency teams feel comfortable.

The original brand safety framework asked a binary question. Is this content harmful. Yes or no. The answer was usually obvious and the action was straightforward.

The brand alignment framework asks a subjective question. Does this environment reflect who we want to be seen as. The answer depends on who is in the room, what the prevailing mood is, and how much courage anyone has to challenge the consensus.

Perhaps the industry would benefit from separating the two questions again. Keep the genuine safety standards. They exist for good reason and they should be rigorous. But recognise that the cultural alignment layer is a different decision entirely, one that carries real commercial cost and deserves to be evaluated on evidence rather than instinct.

Neutrality is not an absence of conviction. It is a disciplined choice to let the numbers lead.

Curious how others see it.