Indie video game publisher Finji has publicly accused social media giant TikTok of using generative artificial intelligence to create and publish unauthorized advertisements for its games, with the AI-generated content reportedly featuring racist and sexist imagery that the publisher could not edit or remove. The allegations, detailed in a series of posts from Finji’s official account, reveal a disturbing new frontier in digital advertising where platforms may be autonomously creating and monetizing content using a company’s intellectual property without consent.
The Unauthorized AI Ad Campaign
According to Finji, the publisher discovered that TikTok had generated and was running advertisements for its critically acclaimed narrative adventure game, Night in the Woods, and its cooperative survival game, Tunic. These ads were not created or approved by Finji’s marketing team. Instead, they were allegedly produced by TikTok’s own AI systems, which scraped game assets and promotional materials to create new video advertisements. The core of the controversy lies not just in the unauthorized use of IP, but in the specific, harmful content the AI produced. Finji stated that the AI-generated ads included “racist and sexist imagery,” fundamentally misrepresenting the games and their creators’ values.
“We do not approve these ads. We can’t edit them. We can’t turn them off. We are being charged for them,” Finji’s statement read. This creates a triple violation: the theft of creative assets, the generation of offensive content, and the financial burden placed on the victim. The publisher emphasized that the ads were not simple edits or template-based promotions but were wholly new creations birthed from generative AI tools operated by the platform itself.
A Broken System of Control and Consent
The incident exposes critical flaws in the advertising interfaces and policies of major social platforms. Finji reported that the ads appeared in their TikTok Ads Manager account, seemingly initiated by the platform’s automated systems. Despite being listed in their account dashboard, the publisher found the ads were locked—options to edit, pause, or delete them were non-functional or grayed out. This lack of control turned their own advertising account into a vehicle for content they despised.
The Financial and Ethical Burden
Perhaps the most galling aspect for the small indie publisher was the financial model. Finji was reportedly being charged for the impressions and clicks generated by these unauthorized, AI-created ads. This means TikTok’s platform allegedly used Finji’s IP to create offensive content, published it under Finji’s account, and then billed Finji for the privilege. This represents a profound breakdown in the fundamental advertiser-platform relationship, where the platform is no longer a neutral host but an active, unapproved creative agent spending the client’s money.
Generative AI’s Propensity for Bias and Harm
The case of Finji provides a concrete, real-world example of the known risks associated with generative AI. Large language and image models are trained on vast datasets scraped from the internet, which inherently contain societal biases, stereotypes, and harmful content. When these models generate new content without careful guardrails and human oversight, they frequently reproduce and amplify these biases. For Finji, this meant their characters and worlds were reinterpreted through an AI lens trained on a corpus likely filled with reductive and offensive tropes about gender and race.
This is not a hypothetical concern. Studies and numerous prior examples have shown AI image generators struggling with race and gender, often defaulting to stereotypes or producing grotesque inaccuracies. When such technology is deployed at scale by a platform like TikTok, operating autonomously on client assets, the potential for brand damage is immense. The AI does not understand context, narrative, or authorial intent; it merely identifies patterns and replicates them, often with socially toxic results.
The Legal and Intellectual Property Quagmire
The situation plunges into a legal gray area. Who owns the copyright to an ad created by TikTok’s AI using Finji’s copyrighted game characters and scenes? Who is liable for the defamatory or harmful nature of the content—the platform that created it, or the advertiser whose account it runs under? Traditional advertising agreements do not account for a scenario where the platform becomes the content creator. Finji’s experience suggests that current Terms of Service may grant platforms broad, ambiguous rights to “optimize” or “enhance” ad campaigns, potentially covering such AI generation in fine print that most advertisers never scrutinize.
Industry-Wide Implications for Digital Advertising
Finji’s ordeal is likely not an isolated incident but a canary in the coal mine for a widespread, opaque practice. Major platforms like TikTok, Meta, and Google are aggressively integrating generative AI into their advertising suites, offering tools for automated ad creation, text variation, and image generation. The promise is efficiency and performance. The hidden risk, as Finji discovered, is a complete loss of brand control and the potential for AI to act in ways that are directly contrary to a company’s ethics and public image.
For the indie game development scene, which often operates on razor-thin margins and relies on a direct, authentic connection with its audience, such AI malfeasance is particularly devastating. A game’s aesthetic and narrative sensibility is its lifeblood. To have that distorted by an AI into something racist or sexist can irrevocably harm a studio’s reputation and relationship with its community. It also raises the question: if this can happen to a visible publisher like Finji, how many smaller developers are experiencing similar issues without the platform or voice to call it out?
The Platform’s Responsibility and Opaque Algorithms
The incident places the onus squarely on TikTok to explain its systems. Advertisers deserve transparency about when and how AI is being used to modify or create content on their behalf. There must be clear, accessible opt-out mechanisms, and human oversight must be a mandatory component of any process that can generate novel content. Furthermore, platforms must bear full financial and legal responsibility for any harm caused by content their own AI systems generate without explicit, informed consent from the advertiser. The current model, where the advertiser is billed and held liable, is untenable.
Broader Questions for the AI-Driven Web
Finji’s struggle transcends advertising and touches on the core anxieties of the generative AI era. As companies integrate these tools into core products, they are outsourcing creative and communicative acts to stochastic systems with no moral compass. This case illustrates the “content laundering” potential of AI: a platform can take a creator’s work, transform it into something objectionable, and present it back to the world under the creator’s name, all while profiting from the engagement. It undermines the very concept of authorship and accountability in the digital sphere.
The response from the tech community and other advertisers will be critical. If Finji’s claims are validated, it could lead to class-action lawsuits, regulatory scrutiny, and a mass advertiser revolt against opaque AI ad tools. Regulators in the EU under the Digital Services Act and in the U.S. may find this a fertile case for examining platform liability and the need for “AI transparency” mandates in commercial tools.
The digital landscape is being reshaped by autonomous AI agents making decisions that affect brands, reputations, and cultural discourse. Finji’s experience is a stark warning that without robust safeguards, clear consent frameworks, and ultimate human accountability, these tools will create crises of trust and cause real harm. The promise of automated efficiency is meaningless if it comes at the cost of a company’s voice, ethics, and control over its own public face. For creators and advertisers everywhere, the message is clear: scrutinize your ad dashboards, demand transparency from platforms, and be prepared to push back against the black box of AI, or risk having your identity rewritten by an algorithm you never asked for and cannot control.
