AI Agent Ads Are Already Breaking the Ad Stack

Roughly 60% of web traffic is now non-human. We’ve known this for years and mostly filed it under ‘bot fraud.’ But this week handed us something different — a named AI system, Perplexity, actively blocking ads that a named publisher, Time, was serving directly into AI agent workflows. The era of non-human audiences with editorial opinions about your creative has arrived.

When the Audience Pushes Back on Your Ads

Perplexity’s decision to block Time’s markdown agent ads — flagged internally as ‘deceptive’ — is the first high-profile instance of an AI intermediary exercising what you might call editorial immunity over paid placements. Digiday broke the story, and while the technical specifics are still emerging, the structural implication is clear: publishers experimenting with agent-native ad formats are running ahead of any agreed protocol, and AI platforms are going to referee that gap themselves.

For programmatic buyers, this creates a new category of inventory risk. You can win the bid, serve the impression, and still have the ad suppressed before it reaches any meaningful decision point in the user journey. In agentic environments — where Perplexity, ChatGPT plugins, and similar systems are increasingly completing tasks on behalf of users — the ad server and the consumption environment are no longer the same thing. That’s not a minor UX wrinkle. That’s a structural breakdown in the impression-to-outcome chain that most measurement frameworks currently assume is intact.

Liftoff’s IPO Argument Is Actually an Ad Stack Argument

While agent ads were getting blocked in New York, Liftoff was on its first earnings call as a public company, making a case that resonates well beyond mobile gaming. CEO Jeremy Bondy’s core message, as AdExchanger reported, was essentially: we are a full-stack mobile performance platform, not a gaming DSP with good PR. Liftoff operates both an SSP (helping apps monetize via SDK) and a DSP (helping advertisers acquire users), which puts it in the same structural conversation as the major walled gardens — just with more transparency about how the pipes connect.

For Southeast Asian growth teams, this matters more than it might for their US counterparts. Mobile app ecosystems in the region — spanning super-apps like Grab, e-commerce platforms like Shopee and Lazada, and their associated ad networks — run on exactly this kind of vertically integrated stack. When a platform controls both supply and demand, bid strategies behave differently. CPMs look efficient right up until you realise you’re optimising inside a closed system with its own definition of ‘conversion.’ Liftoff going public forces that architecture into analyst scrutiny, which should sharpen every buyer’s questions about the stacks they’re already running on.

---

The Leigh Terry Appointment and What Board-Level Media Expertise Signals

Nine Entertainment’s appointment of Leigh Terry — former APAC CEO of IPG Mediabrands, now folded into Omnicom Media — to its board is quieter news, but worth reading as a directional signal. Nine is one of Australia’s largest media companies, and bringing in someone who spent two decades managing media investment across 16 Asia-Pacific markets at scale suggests the board wants someone who understands the buy side from the inside.

The pattern of senior agency executives moving into publisher and platform governance roles is accelerating across the region. It reflects something the ad industry has been slow to say plainly: the line between publisher strategy and advertising strategy is no longer meaningful at the top. Content decisions, inventory architecture, first-party data strategy, and AI readiness are now a single brief. Terry’s hire is Nine hedging on exactly that convergence. For agencies and brand-side media teams in Southeast Asia watching how regional media groups position themselves, it’s a useful read on where board-level priorities are heading.

What This Week’s Signals Mean for Your Next Planning Cycle

Three developments in 48 hours — an AI platform blocking publisher ad innovation, a mobile adtech company reframing its entire market position at IPO, and a veteran media executive crossing from agency to publisher board — aren’t unrelated. They’re three different angles on the same underlying shift: the ad stack is fragmenting again, and the fault lines are forming around AI intermediation, platform verticality, and governance.

The practical implication for media planners and programmatic leads is uncomfortable but actionable. Your current measurement architecture almost certainly doesn’t account for AI agent suppression as a category of non-delivery. Your mobile DSP relationships deserve harder questions about supply-demand integration and what ‘conversion’ means inside their closed loop. And your media mix models need to start treating AI-accessible inventory as a distinct channel with distinct rules — not a subset of display or native.

The brands that get ahead of this won’t necessarily spend more. They’ll ask better questions of their stacks before the next Perplexity-style block surfaces in their own campaign data.

Key Takeaways

• Audit your programmatic measurement framework for AI agent suppression as a distinct non-delivery event — it won’t show up as an error in most current dashboards.
• When evaluating mobile DSPs, map whether they control both supply and demand, and model how that integration affects your effective CPM and attribution logic.
• AI-accessible inventory (agent browsers, LLM plugins, AI-assisted shopping flows) needs its own channel classification in your media mix — the impression dynamics are fundamentally different from traditional digital.

The more interesting question isn’t whether AI agents will become a significant media audience — they already are. It’s whether the industry will build shared protocols for agent-native advertising before every major AI platform starts making unilateral calls about what counts as deceptive. The window for that kind of coordination is probably shorter than the next annual planning cycle.

---

At grzzly, we work with marketing and media teams across Southeast Asia who are navigating exactly this kind of stack complexity — from mobile-first programmatic to emerging AI channel planning. If your measurement framework hasn’t caught up with where your inventory is actually being consumed, that’s a conversation worth having. Let’s talk