AI’s Stress Test: what 2026 means for Publishers and Supply Chains

In the coming year, AI will stress-test all corners of media, and will be stress-tested itself. What comes out the other side is impossible to predict, but we can forecast where the flashpoints of change will erupt.

Search disruption will open new avenues for publishers and brands

AI search has substantial growth potential, and every question answered here is one less click-through to a website.

Fortunately, publishers are well-versed in the capricious nature of referral traffic, and many are already deep into a journey of channel and revenue diversification. Major UK publishers generate further income from subscription revenues, making them less vulnerable to traffic fluctuations. Meanwhile, smaller and independent journalists, especially those that serve specialist or niche audiences, are increasingly forging direct connections with them through multimedia channels such as social media, paid newsletters, and other direct-to-audience platforms.

Front-facing vertical video and podcasts can also be  lucrative growth drivers, reaching audiences where the bulk of attention is now found, while providing an authentic, personality-driven experience that can’t be replicated by AI. The small screen is another example, especially for publishers who produce documentary-style video. Partnerships with broadcasters and streamers, or simply distributing via platforms such as YouTube, can offer a route to CTV expansion, and a safer haven from AI disruption than the open web.

As more consumers research products and, once ChatGPT brings its checkout integrations to the UK, complete purchases entirely within AI interfaces, brands will naturally seek to understand how visibility is earned in these environments.

Generative engine optimisation (GEO) has been positioned as the successor to SEO, but the comparison only goes so far. If deciphering search algorithms was difficult, working out how opaque, constantly evolving large language models (LLMs) surface information will be harder still, especially with chatbots personalising their output to each individual user.

Publishers with licensing arrangements or close relationships with AI companies may find opportunities to influence outcomes through branded content that informs chatbot responses. However, the absence of standardised measurement and attribution will make GEO a moving target. Success will hinge on trial and error, and whether that can be translated into a solid discipline — an area where publishers have long demonstrated strength.

The supply chain will change for the better, but AI might not be why

AI has already become a core component for streamlining workflows, campaign optimisation, and media buying. The evolution toward fully agentic buying and selling, however, will run headlong into the realities of programmatic infrastructure. Real-time bidding operates round the clock, at incomprehensible scale, with transactions measured in milliseconds and little tolerance for error.

Continued advances in infrastructure and AI-native architectures are likely to enable new operating models that extend well beyond today’s API-led workflows. We can expect to see a wave of promising test cases and pilot initiatives as the ecosystem evolves, where agentic systems could become an integral part of programmatic execution. However, the core principles of how media is traded are still a long way from being fundamentally altered.

We’ve seen standardised agentic protocols emerge recently, but translating these blueprints into a live, global supply chain is another challenge entirely. There are still some questions around latency, cost, and the reliability of AI agents at the scale and speed programmatic requires.

Whether or not AI agents live up to the hype, structural changes are coming to programmatic advertising either way. This past year we saw DSPs edge closer to direct supply integrations, while SSPs rolled out frontends for buyers. Both are responding to long-standing demands from either end of the supply chain for simpler, more transparent trading. The likely end point of such convergence is a market dominated by competing end-to-end platforms, where the old acronyms that bookended the supply chain matter less than the ability to offer a lean and auditable path between advertiser and publisher.

This innovation might work in publishers’ favour. Access to quality media and meaningful audience signals have become the most prized assets in digital advertising. With both DSPs and SSPs now courting publishers to strengthen buyer propositions and leverage their consumer proximity, they will have a stronger bargaining position than they have in years. Publishers must utilise this leverage to prevent end-to-end platforms leading to the audience commodification that characterised the ad network era.

2026 will be yet another year of change — and with that, comes a degree of uncertainty. During such times, it’s worth centering on what will always be true: all media, bought or sold, finds its value from its audience. Know your audience  , know their movements, know their preferences and habits, and you can sustain connection and relevance through all coming waves of hype and disruption.