GenAI is disrupting publishing. Advertisers need programmatic AI to help repair it

The relationship between publishing, advertising, and AI is more closely linked than it might first appear. Weakness in one area carries through to the others, writes Onetag’s SVP.

The rapid rise of ChatGPT and Google AI search has forced the media industry into a familiar conversation. How will publishers adapt? What happens to traffic referral and monetisation when AI-generated answers are increasingly blurring the lines between ‘real’ journalism and user-generated content?

These are important questions. But for advertisers, there is a more immediate concern. When the quality of journalism weakens, the effectiveness of the environments it relies on declines with it.

This is as much a performance issue as a publishing issue.

GenAI systems depend on a steady supply of original, diverse, human-created content. Advertising systems depend on the same foundation. When that foundation becomes thinner or less reliable, the impact shows up in both the quality of the AI output and, crucially, in advertiser outcomes. Targeting becomes less precise, and context becomes less meaningful. As a result, performance becomes harder to sustain.

What starts as a shift in how content is produced and discovered quickly feeds into campaign performance.

When journalism weakens, outcomes follow

Journalism takes time, skilled people, and sustained investment. For most publishers, advertising remains the primary way that investment on the open web is funded. Subscriptions and reader revenue contribute, but they rarely cover the full picture, particularly for independent publishers.

As pressure builds on monetisation, the effects are gradual. Resources tighten, and editorial priorities shift, making original reporting harder to sustain. Over time, the depth and diversity of content begins to narrow.

For advertisers, this has direct consequences.

Campaign performance is shaped by the quality of the environments in which ads appear. When those environments become less trustworthy, less distinctive, or less engaging, the signals that campaigns rely on begin to lose clarity. Even the most advanced AI-driven tools struggle when the underlying inputs are inconsistent.

The same applies to GenAI systems. When models are trained on low-quality or repetitive content, their outputs become less useful. The relationship between publishing, advertising, and AI is more closely linked than it might first appear. Weakness in one area carries through to the others.

A fragmented system under pressure

The current state of programmatic advertising has not made this easier.

Over time, the ecosystem has prioritised scale and efficiency, often at the expense of clarity. Expanding supply paths, duplicated inventory, and limited transparency have made it harder to distinguish between high-quality environments and those that simply appear similar on the surface.

Tools designed to manage risk, such as keyword blocking and rigid brand safety frameworks, have added another layer of complexity. While well-intentioned, they often fail to account for context. Responsible journalism can be excluded alongside genuinely unsuitable content, with little visibility into the reasons for those decisions.

At the same time, signal loss has reduced the effectiveness of traditional targeting methods. In some cases, match rates have dropped significantly, leaving large portions of audiences less addressable through conventional approaches. In worst-case scenarios, match rates fell below 30%, meaning 70%+ of users were not matched.

Taken together, these factors have made it more difficult for advertisers to consistently access the environments that support strong outcomes. They have also made it harder for publishers to demonstrate the value of their inventory.

How sell-side AI and curation are restoring value

This is where the role of the sell side is starting to change.

AI-driven programmatic technologies are moving beyond simply passing inventory into the market. They are introducing a layer of intelligence that helps shape supply before it reaches the bidstream.

This is where curation plays a defined role, shaping supply based on context, attention, and performance signals, making higher-quality environments easier to access and act on. Attention metrics, placement characteristics, and live performance data provide a clearer picture of quality at the moment an impression becomes available.

This allows supply paths to become more selective. Environments that consistently deliver engagement and outcomes are prioritised. Those that do not can be filtered out earlier in the process. Rather than broad filtering, this approach is more adaptive, using live data to continuously refine which environments are surfaced and which are deprioritised.

Additionally, higher-quality ad creative can now serve as a live performance signal, with immersive formats that analyse attention and engagement and provide interaction data for further campaign improvement.

For advertisers, this creates more stable performance, with decisions being based on observed behaviour rather than assumptions. For publishers, it creates a clearer link between the quality of their content and the demand it attracts.

As inventory and creative selection become more closely tied to performance signals, this helps reinforce a cycle in which quality content is more consistently recognised, funded, and sustained. Over time, this begins to correct some of the imbalances that have built up in the ecosystem.

Building a more sustainable cycle

There is also a longer-term effect that is easy to overlook.

When high-quality journalism is more consistently recognised and rewarded through advertising investment, it becomes more sustainable. Publishers are better positioned to continue producing the kind of content that supports both advertiser outcomes and AI development.

That, in turn, feeds back into the system. Stronger content leads to better training data for AI models. Better AI outputs improve how users discover and interact with information. More meaningful environments support more effective advertising.

It is a cycle where each part reinforces the others.

The alternative is less encouraging. If quality continues to erode, both AI systems and advertising performance will suffer, with signals becoming less reliable. This will lead to outcomes becoming harder to predict and efficiency gains achieved through automation beginning to stall.

A shared responsibility

AI will continue to evolve. Advertising technology will continue to advance. Neither is slowing down. What matters is how the systems around them are managed.

The connection between journalism, advertising, and AI is not always visible in day-to-day decision-making. But it is fundamental. Advertisers rely on strong environments to deliver results. AI relies on strong content to remain useful. Publishers rely on sustainable economics to continue producing both.

Strengthening one part of that system supports the others.

GenAI is reshaping how content is discovered and consumed. At the same time, programmatic AI is creating an opportunity to rebuild how value flows through the open web. Improving how quality is identified, measured, and rewarded can help ensure that performance remains grounded in something more durable than scale alone.

Elli Papadaki is SVP, global supply at Onetag