Can Anyone Run Winning Ads Now? Deconstructing Tiktok’s Ai Automation Claims
By Marcus Vance
The promise of total automation encountered sharp pushback from performance agencies throughout 2025 and 2026. Buyers grew frustrated with automated black boxes that offered little visibility into where ads actually appeared, how bid ceilings were enforced, or why specific audiences were excluded.
That friction compelled the platform to recalibrate. In January 2026, Social Media Today documented that TikTok added expanded control options to Smart+ campaigns. Instead of forcing marketers into binary choices, engineers introduced modular manual overrides. Media buyers gained back the ability to enforce strict minimum-ROAS floors, place explicit frequency caps, restrict placement exclusions, and deploy custom audience segmentation parameters directly within automated campaign structures.
This course correction confirms an open secret among performance practitioners: unconstrained AI automation favors platform ad inventory utilization, not agency profit margins. When buyers allow systems to run on absolute auto-pilot without constraints, the delivery engine frequently distributes spend across low-intent network placements to hit delivery pacing targets. Regaining granular controls allowed seasoned media buyers to establish guardrails, blending programmatic bidding efficiency with disciplined balance-sheet protection.
Marcus Vance is a cybersecurity auditor and technology writer dedicated to educating the public about online safety, data privacy regulations, enterprise security, and emerging cyber threats.
