Meta Shows the Power of AI Ad Optimization Without New Users
Meta’s latest results underscore how AI ad optimization can dramatically increase revenue without relying on rapid audience growth. The company reported a 33% year-on-year revenue surge, driven largely by its AI-scaled advertising engine. Family daily active people rose only modestly, signalling that the real shift is monetisation efficiency: Meta is using AI to refine targeting, campaign optimisation, pricing efficiency, inventory utilisation, creative recommendations and automated delivery. This allows the platform to serve more relevant ads, improve performance and still raise average ad prices. For advertisers, the upside is better programmatic advertising outcomes with less manual tuning. The downside is that as Meta’s algorithms capture more value from each impression, the platform strengthens its own pricing power. Performance that once depended on human media buyers now resides inside Meta’s opaque AI systems, making comparable results outside its ecosystem increasingly difficult to replicate.
OpenAI Pushes Into Ads With Self-Serve ChatGPT Ads Manager
OpenAI is moving aggressively into AI ad optimization with the expansion of its ChatGPT Ads Manager. The new beta self-serve portal lets advertisers register, set budgets, upload creatives, launch and manage campaigns, and monitor performance in a single interface. Crucially, OpenAI is introducing cost-per-click bidding, giving marketers tighter alignment between spend and user actions generated by their ads. The company expects to generate USD 2.5 billion (approx. RM11.5 billion) in ad revenue this year and is reportedly eyeing USD 100 billion (approx. RM460 billion) by 2030, showing the scale of its ambition. By partnering with major agencies and ad-tech players, OpenAI is embedding generative AI and programmatic advertising workflows into a familiar stack. For brands, that translates into a new, AI-native performance channel where the same audience can be monetised more effectively through conversational placements and predictive bidding strategies.
Predicting Creative Performance Before Launch Becomes a Competitive Edge
Creative is emerging as the next major lever in AI ad optimization, and platforms like Monks’ Creative Intelligence illustrate why. Rather than treating creatives as static files, the system ingests video and static assets, automatically segments them into individual clips and generates rich metadata on visual and audio attributes. These data points are stitched directly into media performance metrics, enabling granular creative performance prediction. Instead of waiting for live results, marketers can understand which sensory elements—such as a specific facial expression, product close-up or brief shot of a red car—are most likely to drive engagement and conversions. With research from Nielsen and the IAB indicating that creative accounts for nearly half of sales impact across channels, automating this analysis is powerful. It helps teams iterate faster, reduce wasteful testing and systematically scale high-performing creative patterns across campaigns and audiences.
Agentic AI Advertising Systems Take Over Optimization Workflows
Agentic AI advertising systems are rapidly becoming standard across performance platforms as marketers chase always-on optimization. Monks.Flow positions itself as an agentic AI system orchestrating marketing workflows, while Taboola’s new research shows the appetite for such automation beyond walled gardens. Agentic AI goes beyond simple algorithmic targeting: it continuously makes decisions on bids, placements, budget allocation and creative rotation with minimal human intervention. According to Taboola’s study, 76% of advertisers are already seeing meaningful uplift from AI-powered solutions in search and social, and 80% say they would immediately increase spend on the open web if comparable agentic tools were available. Yet integration remains a major hurdle, especially for large advertisers managing complex tech stacks. As these systems mature, they promise more efficient programmatic advertising, but they also centralise control of performance levers inside proprietary AI agents.
The New Dependence: Performance Gains and the Risk of Vendor Lock-In
Across Meta, OpenAI and emerging creative intelligence tools, a common pattern is clear: platforms are using AI to extract more value from the same users, while advertisers grow increasingly dependent on their ecosystems. Meta’s AI-driven pricing power, ChatGPT’s cost-per-click bidding and agentic AI systems all shift optimisation from human teams to platform-controlled models. This can deliver superior performance, but it also concentrates knowledge inside closed systems. Advertisers often see fewer levers to pull, reduced transparency into why campaigns work and higher switching costs if they want to diversify away from dominant platforms. Taboola’s findings highlight this tension: brands want agentic AI beyond walled gardens, yet struggle with integration and fragmentation. The next competitive frontier will be balancing the short-term gains of AI ad optimization with strategies to avoid deep vendor lock-in, such as owning first-party data, rigorous experimentation and maintaining cross-platform flexibility.