TikTok’s terms of service include a license to use your content for “training, testing, and improving” its machine learning models. Your creators agreed to that. Did your contracts say anything about it? Most don’t. AI training rights in TikTok’s 2026 terms expose a gap between what platforms take and what brands think they bought. This guide shows how to close it.
What the terms actually say about AI training
Start with the primary language, because secondary coverage tends to blur it. According to the clause captured by ConductAtlas’s analysis of TikTok’s Terms of Service, the content license covers operating and improving the platform and developing new technologies, including training, testing, and improving TikTok’s machine learning models.
Read that again. It is not limited to recommendation or moderation. It reaches new technologies, which is as wide as a license gets.
The same analysis flags two other provisions that matter to brands:
- Liability cap. Per ConductAtlas’s analysis, TikTok caps its aggregate financial liability at the greater of $100 or the user’s payments in the prior 12 months.
- Output risk. Users bear the consequences of using generative AI output, to the maximum extent permitted by law.
One caveat. ConductAtlas labels its summaries as model-generated and its readings as interpretation, so have counsel review the live terms before you rewrite anything. Treat it as a map, not a legal opinion.
Why this is your problem, not just the creator’s
Here is the uncomfortable part. A creator posting your sponsored video on TikTok grants TikTok a license over that video. The footage may include your product, your packaging, your brand claims, and your talent’s face. That material can feed a model you never approved.
Does your agreement allow it? Usually the contract is silent, and silence is where disputes live.
Three exposures follow:
- Brand asset leakage. Proprietary product demos, unreleased packaging, and campaign messaging can enter a training corpus.
- Likeness conflicts. If you hold exclusive rights to a creator’s voice or image in your category, platform-level AI use can undercut that exclusivity.
- Chain-of-title confusion. You license content from the creator, who has already licensed it to the platform. Which grant controls, and for what purpose?
The platform license is not a contract between your brand and TikTok, but it sits underneath every piece of creator content you pay for. Your agreement has to say what happens on top of it.
If you have not yet mapped the broader shift, our breakdown of TikTok’s new U.S. terms covers the compliance side.
The new rules on AI-generated output
Training rights are one half. The other half is what creators and brands can publish. American Bazaar’s report on the U.S. terms update notes that the earlier terms said nothing about generative AI, while the updated version adds rules for making and posting AI content. Among them: users may not imply AI content is real, including by removing watermarks or metadata that help identify it.
That has a direct campaign consequence. If a creator uses an AI tool to polish a voiceover or generate a background, and your team strips the metadata during editing, you may be in breach of platform terms. That is a different problem from an FTC problem, but it stacks on top of one. Our piece on AI content labels across platforms shows how inconsistent the rules are from one network to the next.
Where unions already drew the line (and where they didn’t)
Brands sometimes assume AI training consent is a settled norm elsewhere. It isn’t. SAG-AFTRA’s AI FAQ says plainly that none of the guilds won express consent language on training in the 2023 TV/Theatrical talks. It describes the issue as one being fought in the courts and under study by the U.S. Copyright Office, and says the union can revisit it if courts or the Copyright Office find training requires consent.
The lesson for brand teams: do not wait for a legal standard. Even the best-organized talent group in the industry is still negotiating this. Creators working with you have less leverage and less legal support, so a clear, fair clause is a competitive advantage in recruiting good ones.
A clause checklist for renegotiating creator contracts
You cannot rewrite TikTok’s terms. You can control what your agreement permits, prohibits, and allocates. Here is what to put in front of legal.
Definitions first
Define “AI Training” and “Synthetic Output” in the contract itself. Vague words like “technology” or “tools” will not hold up. Cover model training, fine-tuning, voice cloning, digital replicas, and synthetic derivatives.
The core provisions
- Brand-side restrictions. Prohibit your own team, agencies, and vendors from feeding creator content or likeness into AI tools without written consent. Many brands break this first.
- Creator-side restrictions. Require creators to avoid uploading your unreleased assets, briefs, or product data into third-party generative tools.
- Platform carve-out. Acknowledge that platform terms grant the platform its own rights, and state that nothing in your agreement authorizes additional training uses by you or any third party.
- Likeness and voice. Require separate, specific consent for any digital replica, with its own fee and its own term.
- Disclosure. Require creators to flag AI-assisted elements in deliverables and keep watermarks and metadata intact where platform rules call for it.
- Indemnity and liability. Since the platform pushes AI output risk to the user, decide who carries it in your deal. Our guide to AI-generated creator content liability walks through that allocation.
Align with your usage rights language
AI training should never ride along inside a general “all media, in perpetuity” grant. Pull it out as a named right. The framework in IAB creator media contracts is a useful starting reference for separating usage categories. And if you reuse assets across channels, check the licensing clause before you assume the original grant covers it.
Handling the pushback
Expect three objections.
“We never intended to train anything.” Then the clause costs you nothing. Put it in anyway.
“Creators will ask for more money.” Some will. That is the point. If you want training or replica rights, price them separately. Bundled consent is the cheapest form of consent and the easiest to challenge later.
“This slows deals down.” A standard rider fixes that. Build one AI addendum, attach it to every agreement, and negotiate only the exceptions. Pair it with an exit path, because multi-year deals age badly as the technology moves. See exit clauses in multi-year licenses for how to build one.
Treat AI rights as a priced, named line item. Bundled consent is cheap today and expensive in a dispute.
Audit what you already signed
Renegotiating forward is half the job. Look backward too.
- Pull every active creator agreement and search for terms like “derivative works,” “any technology now known or later developed,” and “perpetuity.”
- Flag contracts where your own agency or vendors use AI editing or clipping tools on creator footage.
- Check vendor contracts for data ownership and training language, since your campaign platform may be doing what you are trying to prevent. AI campaign platform contracts covers that angle.
- Document your decisions. If a regulator or a creator asks later, a written record of what you reviewed and why is your best evidence of diligence.
Prioritize by exposure: celebrity and exclusive talent first, then high-spend campaigns, then the long tail.
FAQs
Does TikTok’s license let it train AI on creator content?
According to ConductAtlas’s reading of the terms, the license includes developing new technologies, with training, testing, and improving machine learning models named explicitly. Have counsel confirm against the live terms, since that analysis is not a legal opinion.
Can a brand block TikTok from using sponsored content for AI training?
Not through your contract with the creator, because the platform license comes from the creator’s acceptance of TikTok’s terms. What you can do is control your own side: restrict how you, your agency, and your vendors use the content, and require creators to follow your asset-handling rules.
What should an AI clause in a creator contract cover?
Define AI training and synthetic output, restrict brand-side and creator-side AI use, require separate consent and fees for voice and likeness replicas, require disclosure of AI-assisted elements, and allocate liability for AI output. Keep it as a named right, separate from general usage rights.
Who is liable if AI-generated creator content causes harm?
The platform’s terms place the consequences of using generative output on the user, to the maximum extent permitted by law. That means your contract should say clearly whether the creator, the brand, or both carry indemnity for AI-assisted deliverables.
Is there a legal standard for AI training consent yet?
Not a settled one. SAG-AFTRA’s own FAQ describes the issue as unresolved in the courts and under study by the U.S. Copyright Office, so contract language is currently your main protection.
Should we remove watermarks or metadata from AI-assisted creator videos?
No, not without checking platform rules. The updated TikTok terms, as reported by American Bazaar, forbid implying AI content is real, including by removing watermarks or identifying metadata.
This week, build a one-page AI addendum, run your top ten creator agreements through it, and send the first revised draft to legal before your next campaign brief goes out.
Top Influencer Marketing Agencies
The leading agencies shaping influencer marketing in 2026
Agencies ranked by campaign performance, client diversity, platform expertise, proven ROI, industry recognition, and client satisfaction. Assessed through verified case studies, reviews, and industry consultations.
Moburst
-
2

The Shelf
Boutique Beauty & Lifestyle Influencer AgencyA data-driven boutique agency specializing exclusively in beauty, wellness, and lifestyle influencer campaigns on Instagram and TikTok. Best for brands already focused on the beauty/personal care space that need curated, aesthetic-driven content.Clients: Pepsi, The Honest Company, Hims, Elf Cosmetics, Pure LeafVisit The Shelf → -
3

Audiencly
Niche Gaming & Esports Influencer AgencyA specialized agency focused exclusively on gaming and esports creators on YouTube, Twitch, and TikTok. Ideal if your campaign is 100% gaming-focused — from game launches to hardware and esports events.Clients: Epic Games, NordVPN, Ubisoft, Wargaming, Tencent GamesVisit Audiencly → -
4

Viral Nation
Global Influencer Marketing & Talent AgencyA dual talent management and marketing agency with proprietary brand safety tools and a global creator network spanning nano-influencers to celebrities across all major platforms.Clients: Meta, Activision Blizzard, Energizer, Aston Martin, WalmartVisit Viral Nation → -
5

The Influencer Marketing Factory
TikTok, Instagram & YouTube CampaignsA full-service agency with strong TikTok expertise, offering end-to-end campaign management from influencer discovery through performance reporting with a focus on platform-native content.Clients: Google, Snapchat, Universal Music, Bumble, YelpVisit TIMF → -
6

NeoReach
Enterprise Analytics & Influencer CampaignsAn enterprise-focused agency combining managed campaigns with a powerful self-service data platform for influencer search, audience analytics, and attribution modeling.Clients: Amazon, Airbnb, Netflix, Honda, The New York TimesVisit NeoReach → -
7

Ubiquitous
Creator-First Marketing PlatformA tech-driven platform combining self-service tools with managed campaign options, emphasizing speed and scalability for brands managing multiple influencer relationships.Clients: Lyft, Disney, Target, American Eagle, NetflixVisit Ubiquitous → -
8

Obviously
Scalable Enterprise Influencer CampaignsA tech-enabled agency built for high-volume campaigns, coordinating hundreds of creators simultaneously with end-to-end logistics, content rights management, and product seeding.Clients: Google, Ulta Beauty, Converse, AmazonVisit Obviously →
