Campaign Asia has published an analysis arguing that artificial intelligence still cannot solve advertising's long-running measurement problem, despite the volume of AI-branded measurement tools now on the market.

The piece addresses a question that media buyers and brand marketers have asked vendors repeatedly over the past two years: if AI can write copy and generate video, why can it not tell an advertiser which half of its budget worked? Campaign Asia's answer, per the headline and framing of the piece, is blunt — the measurement puzzle remains unsolved.

Why does the claim matter?

Ad measurement has been under strain for more than a decade. The industry lost its cross-platform currency as audiences fragmented across search, social, streaming and retail media. Privacy changes on the open web and mobile restricted the identifiers that older attribution models depended on.

Into that gap, measurement vendors have shipped a steady stream of products branded as AI: automated attribution, synthetic panels, predictive audiences and machine-learning brand-lift estimates. Campaign Asia's argument cuts against that marketing push. The publication's core position is that these tools have not yet closed the measurement gap, and that buyers should treat AI claims from vendors with caution.

What does the analysis actually say AI can't do?

The title frames the limitation clearly: AI cannot yet solve the measurement puzzle. That single word — yet — carries weight. It positions the shortfall as current-state, not permanent. But it also signals that the industry's present toolset, whatever the sales decks say, does not deliver a definitive answer on advertising effectiveness.

For operators, the practical takeaway is straightforward:

• AI-branded measurement products remain works in progress rather than settled currencies.
• Marketers should keep demanding evidence behind vendor accuracy claims.
• The structural problems — fragmented audiences and restricted identifiers — predate AI and constrain what any model can do.
• "Not yet" means budget owners should not abandon existing measurement discipline in favor of AI outputs alone.

Who should read this?

The analysis targets the advertising and media community across the Asia-Pacific region that Campaign Asia serves: brand-side marketers, agency planners and measurement leads at publishers and platforms. Anyone currently evaluating an AI attribution or brand-lift product will find the piece a useful counterweight to vendor positioning.

The debate it joins is not academic. Measurement decisions determine budget allocation across channels, and every quarter that passes without a reliable cross-channel answer leaves money parked in the wrong places. A publication of Campaign Asia's standing calling out AI's current limits gives buyers a citable reference point in vendor negotiations.

What happens next?

The word "yet" in the headline leaves the door open. AI capabilities in measurement continue to evolve, and the piece does not argue the tools will never work — only that they have not worked so far. Marketers should expect this argument to sharpen as vendors push back and as more case data accumulates on both sides.

For now, the operating assumption for ad buyers is unchanged: no AI system on the market today fully solves advertising measurement, and any vendor claiming otherwise has a case to prove.