I have been attending Cannes Lions every year since 2022, not to chase the festival hype, but to take a step back and gain a comprehensive view of how global shifts in technology and transaction flows introduce values and specific operational challenges (risks) to the Japanese media and advertising landscape. Beyond merely identifying risks, the true purpose of being on the ground is to discover the new values and hidden investment opportunities embedded within these drastic changes.

The structural shifts observed at last year’s festival—the transition into the deployment phase of "Agentic AI" and the critical warnings surrounding collapsing search referral traffic—have fully materialized over the past 12 months, developing into unavoidable operational challenges, but also unfolding as remarkable new possibilities.

Today, the advertising ecosystem has moved entirely into an algorithmic execution layer. We have progressed beyond single, isolated agents working end-to-end (the 3rd wave). The industry is now transitioning into the 4th wave: "Swarms of Agents" running continuously on cloud infrastructure, automatically picking up tasks via webhooks and system triggers to execute end-to-end operations without human oversight.

As consumers bypass traditional browsing to rely entirely on these conversational AI swarms for purchasing decisions—a phenomenon known as Agentic Commerce—we face a next-generation structural challenge: Must we design our advertising for human consumers, or for AI agents? Consequently, navigating "Machine Attention"—optimizing content to be consumed and evaluated by background AI agents—has become the new battleground of marketing.

When evaluating the trajectory of the publishing and advertising industries toward AGI (Artificial General Intelligence), the following milestones present a realistic roadmap calculated backward from current technical evolutionary speeds and active protocol standardizations across global tech communities.

• Milestone 1 (Near-term): Standardization of Agent-to-Agent (A2A) Commerce Protocols and Governance
  
  A new ecosystem will stabilize where buy-side purchasing agents and sell-side yield-maximization agents negotiate the value, context, and brand suitability of ad inventory in milliseconds without human intervention. Looking ahead to the 5th wave—where swarms utilize end-to-end Reinforcement Learning (RL) to train successive generations of agents based on transaction outcomes—addressing the measurement crisis and establishing legal frameworks like AI accountability will become the primary operational focus.
  
• Milestone 2 (Mid-term): Dynamic Ad Generation via Edge AI and Privacy Sovereignty
  
  Highly efficient, low-power LLM models will be natively integrated into user devices. By processing consumer emotions and immediate context locally without relying on the cloud, devices will dynamically generate personalized, conversational, and shoppable ad experiences in real time. This represents a critical technical breakthrough aligned with tightening digital and data sovereignty regulations.
• Milestone 3 (The AGI Era): "Social Trust" and the Establishment of "Data Licensing" as Ultimate Scarce Resources In a world saturated with AI-summarized, synthesized information generated by relentless agent swarms, the value of premium media—the entities holding verified, primary data and deeply rooted social trust—will paradoxically skyrocket. Advertising models dependent on sheer traffic volume will vanish. Instead, unified publishers retaining absolute data sovereignty are projected to establish robust "data licensing" frameworks, enforcing direct licensing contracts and pro-rata attribution payouts from AI platforms.

To maintain this clear macro perspective of these shifts and convert hidden opportunities into actual growth, we must evaluate operational risks with transparency, identifying exactly which layer requires pragmatic intervention right now.

An objective macro analysis confirms that the Japanese market remains heavily indexed on defensive measures—such as legal frameworks and crawler blocking—lagging behind the aggressive commercialization executed in the West. However, from a strict strategic standpoint, this reactive positioning can be converted into a highly calculated second-mover advantage. By maintaining a deliberate step back, we are positioned to post-analyze the systemic risks—such as short-selling data sovereignty or exhausting operational efficiency—currently penalizing early adopters.

In a marketplace where optimization engines and autonomous agent swarms programmatically buy inventory on their own terms, publishers run the risk of having their identity and contextual nuances reduced to basic variables inside someone else's platform. However, if publishers generate and manage highly precise signals—such as context, sentiment, and attention scores—and feed them directly to buy-side AI models and DSPs, this risk transforms into an entirely new competitive advantage and a substantial avenue for growth independent of walled platforms.

Leveraging this second-mover latency means we must focus our execution strictly on capturing the absolute upstream of the supply chain through resilient infrastructural engineering, rather than conceptual ideals. While taking a step back to analyze the evolution of this era, we remain fundamentally optimistic about a future where a media’s core asset—social trust—properly functions and continuously generates new value.