What changed in Google Performance Max auto-generated asset disclosure in the May 2026 update?

Google Performance Max — Google's goal-based campaign type that combines Search, Display, YouTube, Discover, Gmail, and Maps inventory under a single AI-driven optimisation layer — has been progressively tightening its auto-generated asset disclosure framework since 2024. The May 2026 update is the third substantive iteration and operationalises several disclosure obligations that converge with the EU AI Act, the FTC's AI endorsement guidance, and several state-level AI disclosure frameworks. The update introduces three substantive changes that affect every Performance Max campaign. First, the auto-generated asset labelling is now visible to consumers rather than only available in Google Ads transparency tools. Variants created by Google's AI from advertiser-supplied source assets now carry a visible AI-generated label in placements where the labelling is required by applicable regulation, including EU surfaces under the AI Act, California surfaces under AB 3211, and several other jurisdictions. The labelling appears in the ad creative itself rather than in metadata that consumers cannot see. Second, the asset approval cap is narrowed. Performance Max previously allowed Google's AI to generate and serve auto-variants without prior advertiser approval, with the advertiser able to review and exclude variants after the fact. The May 2026 update introduces an explicit cap on the proportion of auto-generated assets that can serve without advertiser review — currently set at thirty percent of total impression delivery for the campaign. Above the cap the advertiser must approve specific auto-variants before they enter delivery rotation. Third, the advertiser-side review burden is operationalised through a new asset review interface within Google Ads. Advertisers can review proposed auto-variants in batches, approve or reject individual variants, and configure standing rules for variant types they pre-approve. The new interface replaces the previous flow where advertisers reviewed variants after delivery began. From the advertiser perspective the May 2026 update produces several specific operational requirements. Brand-creative-review processes that previously operated on advertiser-supplied source assets now must extend to AI-generated variants. Compliance review processes must verify that the AI-generated variants align with regulatory requirements including the EU AI Act, FTC guidance, and applicable state-level frameworks. Documentation of the review process supports response to regulatory inquiries. For the broader Google Ads policy framework, see Google Ads Policy Guide and track in-flight platform updates through the Policy Tracker .

How do the new visible AI-generated labels work in Performance Max placements in 2026?

The visible AI-generated labels in Performance Max placements operationalise a multi-jurisdiction labelling obligation that converges several regulatory frameworks. The labelling is calibrated to satisfy the EU AI Act Article 50 transparency obligations, California AB 3211 synthetic content provisions, and several other state-level frameworks while remaining commercially functional within Google's ad placement design. The labelling format varies by surface. On Search and Display surfaces the label appears as a small text indicator adjacent to the ad creative reading AI-generated or Generated by Google AI depending on the variant origin. The label is sized and positioned to be clearly visible without overwhelming the ad creative. On YouTube placements the label appears as an overlay during the first three seconds of video content and remains accessible through the ad's information panel. Audio variants on YouTube and Discover include an audio disclosure within the first three seconds. On Gmail and Discover surfaces the label appears within the ad card's metadata area in a position that is visible without expanding the card. On Maps placements the label appears within the location card in an information area that surfaces alongside other content metadata. The labelling is applied automatically by Google when the variant is auto-generated. Advertisers do not need to manually add the label to source creative — the label is added at delivery time based on the variant's origin. The label is removed when the variant is replaced by an advertiser-supplied asset that does not require labelling. The labelling is applied with regional sensitivity. EU surfaces under the AI Act receive labelling for a broader set of variant types including AI-generated text, AI-edited imagery, AI-synthesised voice, and AI-generated video. US surfaces receive labelling primarily for AI-generated imagery and AI-synthesised voice content. Other regions receive labelling based on applicable local frameworks. From the advertiser perspective the labelling does not require source creative changes but does require attention to several specific scenarios. Source creative that the advertiser supplies is not labelled even if the source itself was AI-generated outside Google's system. The advertiser bears the FTC and state-level disclosure obligation for source creative independent of Google's labelling. Performance metrics may differ across labelled and unlabelled variants. Advertisers should monitor performance differentials and adjust creative supply accordingly. Brand-safety incidents involving labelled variants typically land within the advertiser's brand-safety responsibility regardless of the AI generation. The advertiser cannot offload brand-safety liability to Google's AI generation step. For automated review of creative configurations, route through AI Compliance Audit .

How does the asset approval cap work and what advertiser workflow does it require in 2026?

The thirty-percent asset approval cap that the May 2026 update introduces is the most operationally significant change for Performance Max advertisers because it shifts the advertiser-side workflow from post-delivery review to pre-delivery approval for a meaningful portion of variants. The cap operates as follows. Performance Max continues to auto-generate variants from advertiser-supplied source assets, but the proportion of auto-generated variants that can serve without explicit advertiser approval is capped at thirty percent of total impression delivery for the campaign measured over a rolling thirty-day window. Above the cap the advertiser must approve specific auto-variants before they enter delivery rotation. Variants below the cap continue to operate under the previous post-delivery review model. The cap is calculated on impression delivery rather than on variant count because impression delivery captures the actual exposure to consumers. A high-performing auto-variant that delivers a large share of impressions counts more toward the cap than a low-performing variant that delivers few impressions. Approval workflow operates through a new asset review interface within Google Ads. Advertisers receive notifications when proposed variants require approval, with batch review supported for efficiency. Each variant displays the source assets used, the AI-applied transformations, the projected performance based on Google's historical data for similar variants, and any flagged compliance concerns. The advertiser can approve, reject, or modify the proposed variant. Standing approval rules allow advertisers to pre-approve specific variant types — for example, all auto-cropping variants or all colour grading variants — without per-variant review. The standing rules reduce review burden for low-risk variant types while preserving advertiser control over higher-risk types. From the advertiser perspective the cap requires several specific operational changes. The brand-creative-review process must extend to auto-generated variants requiring approval. Review staffing must be sized to handle the variant flow without bottlenecking campaign delivery. Standing approval rules must be configured deliberately to balance review burden and creative control. Documentation of the approval process supports response to regulatory inquiries and provides evidence in performance disputes. The cap creates strategic implications for creative supply. Advertisers supplying a rich source-creative library reduce the AI-generation pressure and stay below the cap with less approval burden. Advertisers supplying minimal source creative push more variant generation through the AI layer and face higher approval burden. The strategic incentive is to invest in source-creative production. The cap also affects creative testing strategy. Advertisers running A/B tests on creative variants should configure the test to avoid hitting the cap, which would produce non-uniform delivery patterns. Testing protocols should be reviewed against the cap to ensure consistent test conditions. For automated review of creative configurations and approval workflow, run AI Compliance Audit .

How does Performance Max AI disclosure interact with FTC guidance, EU AI Act and state-level frameworks?

Performance Max AI disclosure operates within a multi-jurisdiction regulatory framework that includes the FTC's AI endorsement guidance, the EU AI Act Article 50, California's AB 3211, and several other state-level and national frameworks. The platform-side disclosure is calibrated to satisfy the union of applicable requirements but advertiser-side compliance still requires independent attention. The first interaction is with the FTC's May 2026 AI endorsement guidance. The FTC guidance applies to AI-generated content used in commercial endorsement contexts. Performance Max ads typically constitute commercial endorsement when the ad promotes a specific product or service. The platform-side AI labelling satisfies the FTC's transparency requirement at the platform level, but advertisers remain liable for source-creative AI generation that occurs before assets enter Performance Max. The FTC's per-violation penalty ceiling of fifty-three thousand and eighty-eight dollars applies to AI endorsement violations. The second interaction is with the EU AI Act Article 50. The Act requires that interactions with AI systems be transparent to users and that AI-generated content be identified as such. Performance Max ads served on EU surfaces fall within Article 50 when the variants are AI-generated. The platform-side labelling satisfies Article 50 at the placement level. Advertiser-side compliance includes ensuring that source creative supplied to Performance Max also satisfies Article 50 where applicable. The third interaction is with California's AB 3211 synthetic content provisions. AB 3211 requires platform-level labelling and watermarking of AI-generated content used commercially. Google's platform-side labelling satisfies the AB 3211 platform requirement. Advertiser-side compliance includes alignment of source creative with AB 3211's broader synthetic content framework. The fourth interaction is with several state-level frameworks including New York's synthetic performer law, Tennessee's ELVIS Act for voice cloning, and Texas's AI disclosure provisions. The platform-side labelling generally satisfies the state-level platform obligations. Advertiser-side compliance varies by state framework and may extend beyond what the platform-side labelling addresses. The fifth interaction is with applicable industry-specific frameworks. Pharmaceutical advertising, financial services advertising, regulated industry advertising, and other sector-specific frameworks impose additional disclosure obligations that operate alongside the AI disclosure framework. Advertisers running campaigns in regulated industries face the union of AI disclosure plus sector-specific framework obligations. From the operational perspective Performance Max advertisers should not assume that platform-side AI labelling fully satisfies the advertiser's compliance posture. The platform handles placement-level disclosure but the advertiser bears responsibility for source creative, sector-specific framework compliance, and several adjacent obligations. Advertisers running large-scale Performance Max campaigns should commission a compliance review against the multi-jurisdiction framework and document the review for response to regulator inquiries. For consolidated regulatory framework, see FTC AI Endorsement Rule Update and run AI Compliance Audit .

What practical workflow should advertisers follow to operationalise the May 2026 Performance Max changes?

The practical workflow for advertisers to operationalise the May 2026 Performance Max changes involves five workstreams that should run during the rollout period. The workstreams operationalise the AI labelling visibility, the asset approval cap, the advertiser-side review interface, and the multi-jurisdiction regulatory framework. The first workstream is creative supply audit. Advertisers should audit their source-creative library to identify gaps that will produce elevated AI-generation pressure under the cap. Source-creative variants that the advertiser supplies reduce the AI-generation pressure and the approval burden. Creative supply audit produces a remediation plan for source-creative production. The second workstream is approval workflow setup. The new asset review interface within Google Ads should be configured with appropriate access controls, notification settings, and standing approval rules. Standing rules should be configured deliberately to balance review burden and creative control. Standing rules that pre-approve too broadly transfer creative-control responsibility to Google's AI in ways that may not align with brand standards. Standing rules that approve too narrowly create review bottlenecks that delay campaign delivery. The third workstream is brand-creative-review process update. Existing brand-creative-review processes designed for advertiser-supplied source assets must extend to AI-generated variants requiring approval. The review process should include compliance verification against the EU AI Act, FTC guidance, and applicable state-level frameworks. Review staff should be trained on the new variant types and the AI-specific compliance requirements. The fourth workstream is documentation and audit trail. The approval process produces documentation that supports response to regulatory inquiries and provides evidence in performance disputes. Documentation should include the variant approved, the source assets used, the AI transformations applied, the compliance review outcome, and the approval timestamp. Documentation should be retained for at least seven years in line with general advertising record retention. The fifth workstream is performance monitoring against the cap. Advertisers should monitor the proportion of impressions delivered through auto-generated vs advertiser-approved variants and adjust source-creative supply accordingly. Hitting the cap produces non-uniform delivery patterns that can compromise campaign performance and creative testing. Performance monitoring should be integrated into the standard campaign management cadence rather than treated as a separate workstream. From the strategic perspective the May 2026 update reframes Performance Max from a fully-automated optimisation layer to a hybrid optimisation layer with advertiser-side approval for a meaningful portion of variants. Advertisers running large-scale Performance Max programmes should treat the changes as an opportunity to invest in source-creative quality and advertiser-side creative control rather than as a compliance burden. The strategic shift produces better creative output and better regulatory alignment. For end-to-end audit of Performance Max campaign operations, run AI Compliance Audit and reference the broader Google Ads framework through Google Ads Policy Guide .

Google Performance Max AI Disclosure May 2026: Asset Guide

Google Performance Max Auto-Generated Asset Disclosure May 2026: AI Creative Labelling, Asset Approval Cap & Advertiser Workflow

Inside This Compliance Report

1Performance Max May 2026 Update Overview
2Visible AI-Generated Labels
3Asset Approval Cap & Advertiser Workflow
4FTC, AI Act & State-Level Interaction
5Practical Operationalisation Workflow
6Performance Max Compliance Checklist
7Frequently Asked Questions

Performance Max May 2026 Update Overview

Performance Max — Google's goal-based campaign type combining Search, Display, YouTube, Discover, Gmail, and Maps inventory under a single AI-driven optimisation layer — has been progressively tightening auto-generated asset disclosure since 2024. The May 2026 update is the third substantive iteration and converges several regulatory frameworks: the EU AI Act, FTC AI endorsement guidance, and state-level AI disclosure frameworks.

Three substantive changes ship in May 2026. Auto-generated asset labelling is now visible to consumers rather than only available in transparency tools. The asset approval cap is narrowed — auto-generated assets cannot serve more than thirty percent of total impression delivery without advertiser approval. The advertiser-side review burden is operationalised through a new asset review interface within Google Ads with batch review and standing approval rules.

From the advertiser perspective the changes shift Performance Max from a fully-automated optimisation layer to a hybrid layer with explicit advertiser-side approval for a meaningful portion of variants.

"Performance Max is no longer a black box. The May 2026 update brings advertiser approval back into the workflow for the variants that matter, and consumers see the AI-generation label."
— AuditSocials Performance Max brief, May 2026

For the broader Google Ads policy framework, see Google Ads Policy Guide. Track in-flight platform updates through the Policy Tracker.

Visible AI-Generated Labels

Labelling is calibrated to satisfy the union of regulatory requirements (EU AI Act Article 50, California AB 3211, state-level frameworks) while remaining commercially functional within Google's ad placement design.

Surface-Specific Label Format

Surface	Label Format	Position
Search & Display	"AI-generated" / "Generated by Google AI" text indicator	Adjacent to ad creative
YouTube	Overlay first 3 seconds + info panel	Persistent in info panel
Audio (YouTube/Discover)	Audio disclosure first 3 seconds	Pre-creative
Gmail / Discover	Card metadata area indicator	Visible without expanding
Maps	Location card information area	Alongside content metadata

Regional Sensitivity

EU (AI Act): Labelling for AI-generated text, edited imagery, synthesised voice, generated video
US (FTC + state-level): Labelling primarily for AI-generated imagery and synthesised voice
Other regions: Labelling based on applicable local frameworks

Advertisers do not add labels manually — applied automatically at delivery time. Source creative supplied by advertiser is NOT labelled even if AI-generated externally; advertiser bears independent disclosure obligation. For automated review of creative configurations, route through AI Compliance Audit.

Asset Approval Cap & Advertiser Workflow

The thirty-percent cap is the most operationally significant change. Auto-generated variants cannot serve more than 30% of total impression delivery (rolling 30-day window) without explicit advertiser approval.

Cap Mechanics

Calculation basis: Impression delivery, not variant count (high-performers count more)
Below cap: Post-delivery review (previous model)
Above cap: Pre-delivery approval required
Window: Rolling 30-day measurement

Approval Workflow Components

Component	Function
Asset review interface	Batch review of proposed variants with source + transformations + projected performance + flagged compliance concerns
Standing approval rules	Pre-approve variant types (auto-cropping, colour grading) without per-variant review
Notification system	Triggers when proposed variants require approval
Documentation trail	Audit log for regulatory inquiry response

Strategic Implications

Rich source library: Reduces AI-generation pressure, lower approval burden
Minimal source library: Pushes more variants through AI, higher approval burden
Creative testing: Configure tests to avoid cap (non-uniform delivery compromises tests)

For automated review of creative configurations and approval workflow, run AI Compliance Audit.

FTC, AI Act & State-Level Interaction

Platform-side labelling is calibrated to satisfy the union of frameworks, but advertiser-side compliance still requires independent attention for source creative and sector-specific obligations.

Adjacent Framework Mapping

Framework	Platform-side coverage	Advertiser-side residual
FTC AI endorsement guidance (May 2026)	Placement-level transparency	Source creative AI disclosure; $53,088 per-violation penalty
EU AI Act Article 50	EU surface labelling	Source creative compliance; sector-specific obligations
California AB 3211	Platform-level labelling + watermarking	Source creative alignment with broader synthetic content framework
NY synthetic performer / TN ELVIS Act / TX AI disclosure	Platform-level coverage	State-specific creator/cloning consent
Sector-specific (pharma, financial services)	Not covered	Full sector compliance + AI disclosure stack

For consolidated regulatory framework, see FTC AI Endorsement Rule Update.

Practical Operationalisation Workflow

Five-workstream rollout during the May 2026 transition period.

Workstream Summary

Workstream	Output
Creative supply audit	Source-creative gaps + production remediation plan
Approval workflow setup	Access controls, notification settings, standing approval rules
Brand-creative-review process update	Extended to AI variants; compliance verification against AI Act/FTC/state
Documentation + audit trail	7-year retention of variant + source + transformations + compliance review
Performance monitoring against cap	Auto-generated vs approved variant impression share monitoring

For end-to-end audit, run AI Compliance Audit.

Performance Max Compliance Checklist

[ ] Audit source-creative library for gaps that elevate AI-generation pressure
[ ] Configure asset review interface access + notification settings
[ ] Set standing approval rules deliberately (balance burden vs creative control)
[ ] Extend brand-creative-review process to AI-generated variants
[ ] Train review staff on new variant types and AI-specific compliance
[ ] Document approval decisions with source + transformations + compliance review
[ ] Retain documentation for 7 years
[ ] Monitor impression share of auto-generated vs advertiser-approved variants
[ ] Configure creative tests to avoid hitting the cap (non-uniform delivery breaks tests)
[ ] Verify source creative AI compliance independent of platform-side labelling
[ ] Pre-clear regulated-industry campaigns through legal + product review
[ ] Track in-flight Performance Max + AI Act + FTC guidance through the Policy Tracker

Don't miss the next policy change.

Subscribe to the Policy Tracker — get weekly digests or instant Pro alerts across all 8 platforms. Or try our free Keyword Risk Checker first.

Subscribe Free

Report Keywords — Run AI Compliance Audit

#Google Ads#Performance Max#AI Creative#AI Disclosure#Auto-Generated Assets#Ad Compliance#AI Labelling#EU AI Act#FTC#2026 Policy#Advertisers#Compliance Guide 2026

Share This Report

Tweet Share

Google Demand Gen Campaign Compliance Guide 2026: Creative Requirements, Audience Signals & Policy Differences from Discovery Ads

Google's Demand Gen campaigns replace Discovery Ads with expanded placements, AI creative tools, and lookalike segments — but compliance requirements differ significantly across YouTube, Discover, and Gmail surfaces.

2026 US Midterm Political Ad Blackouts: Meta, Google & TikTok Calendar, Pre-Clearance Windows & Advertiser Workflow

The 2026 US midterms move major-platform political ad rules from theoretical to operational. Meta runs an election-week blackout, TikTok maintains its absolute ban, Google enforces pre-clearance windows — and non-political advertisers feel the spillover.

EU DSA Second-Wave VLOP Designations April 2026 — 12+ New Platforms Under Article 33, Cross-Product User Counts & 2027 Audit Timeline

The European Commission's second-wave DSA designations effective April 2026 add 12+ platforms to the Very Large Online Platform list under tighter user-count methodology. The January 2027 compliance review will be the first formal audit of the second-wave cohort with fines up to 6% of global revenue.

Google Performance Max Auto-Generated Asset Disclosure May 2026: AI Creative Labelling, Asset Approval Cap & Advertiser Workflow