Google Ads AI Video Voice-Over: Complete Guide to the New Feature (2026)

Introduction

Google Ads is deepening its AI integration with two major video ad features: an AI voice-over enhancement that automatically narrates videos using existing text assets, and Veo, a generative video model that turns images into high-quality clips. Both tools aim to reduce production friction for advertisers while increasing creative output. The voice-over feature, first reported in March 2026 and expanded in June 2026, is now available as an opt-out enhancement for Performance Max and Demand Gen campaigns.

What Is the Google Ads AI Voice-Over Feature?

The Google Ads AI voice-over feature is a machine-learning-powered tool that generates a spoken narration track for video ads that lack an original voice recording. It selects relevant text assets from the ad group (such as headlines, descriptions, or call-to-action phrases) and uses text-to-speech AI to create a natural-sounding voiceover. According to Search Engine Land, the rollout began in March 2026 as an opt-out addition to Performance Max campaigns, meaning the AI would automatically enhance eligible videos unless advertisers explicitly disabled it.

The feature is opt-out by design, which has drawn both praise and concern. Advertisers who want full control over their ad audio must navigate the campaign-level settings to turn it off. Google provides controls at the campaign level, as detailed in its support documentation, allowing advertisers to enable or disable the voice-over enhancement, along with other video enhancements like automatic resizing and short version creation.

How Does the AI Voice-Over Work in Performance Max and Demand Gen?

The key change is that Google Ads now applies AI voice-overs automatically to video assets that lack an existing voice track, leveraging text from the ad's messaging. A PPC Newsfeed report from June 2026 notes that the Voiceover Video Enhancement is available for both Performance Max and Demand Gen campaigns. The AI synthesizes speech from the advertiser's text assets, then layers it onto the video. The resulting voice-over can be tested via A/B experiments, and performance data (such as view-through rate or conversion lift) is surfaced in the campaign reporting.

Google's advanced reporting and controls, documented in a separate support page, allow advertisers to see exactly which video assets have been enhanced with AI voice-overs, toggle the feature on or off, and run creative experiments to compare enhanced vs. original versions. This granularity is critical for advertisers who want to maintain brand voice consistency while still benefiting from automation.

Google Ads Video Enhancements: Voice-Over, Resizing, and Short Versions

The voice-over feature is part of a broader suite of video enhancements Google offers. The official Google Ads Help center outlines three main automation tools:

• AI voice-over: Adds a narrated track to videos without existing audio.
• Automatic resizing: Adapts video assets to different aspect ratios (e.g., square for YouTube Shorts, vertical for mobile feeds).
• Short version creation: Generates truncated clips from longer videos.

All enhancements are automated but can be controlled at the campaign level. Advertisers using Performance Max or Demand Gen campaigns will see these options under the “Video enhancements” settings. The voice-over feature, specifically, requires that the campaign have text assets (headlines, descriptions) that Google can synthesize into speech.

Advanced Reporting and Controls for AI Video Enhancements

To address advertiser concerns about losing creative control, Google has introduced campaign-level reporting and controls for automated video assets. In the advanced reporting documentation, Google explains that advertisers can:

• View enhancement status: See which videos have been enhanced with AI voice-overs, resizing, or short versions.
• Run A/B experiments: Compare the performance of enhanced vs. unenhanced video assets within the same campaign.
• Toggle enhancements on/off: Enable or disable each enhancement type individually.
• Download performance reports: Export data on impressions, clicks, and conversions for enhanced assets.

These controls are a direct response to earlier feedback that AI enhancements should be transparent and reversible. By placing the power in advertisers' hands, Google hopes to increase adoption without alienating brands that rely on highly produced creative.

Veo in Google Ads: Generative Video from Images

Beyond voice-overs, Google is pushing generative video creation through its Veo model. Announced on the Google Business blog, Veo is now available in Asset Studio within Google Ads. It enables advertisers to turn static images into high-quality video clips, reimagine existing assets, and generate multiple variations for different YouTube audiences.

Veo works by taking an image input and producing a short video that reflects the visual style and content of the original image. For example, a product photo could be transformed into a lifestyle video scene. This feature is still in early stages and may require a Google Ads premium tier or additional approvals. According to the announcement, Veo aims to help advertisers “resonate with the YouTube audiences that matter most” by creating video content that performs well across formats.

AI Voice-Over vs. Veo: Comparison Table

Below is a comparison of the two AI-driven video features currently available in Google Ads:

Feature	AI Voice-Over	Veo Generative Video
Input	Existing text assets + video without audio	Static images
Output	Narrated video (same visuals, new audio)	New video clips (new visuals, with or without audio)
Where available	Performance Max, Demand Gen campaigns	Asset Studio within Google Ads
Opt-in/Opt-out	Opt-out (on by default); can disable at campaign level	Requires manual creation in Asset Studio
Use case	Faster production of audio for existing video	Creating video from still images when no video exists
Reporting	Campaign-level reporting, A/B experiments	Creative performance data via Asset Studio
Announcement date	March 2026 (voice-over); June 2026 (expansion to Demand Gen)	Announced alongside AI voice-over; available within months

Practical Implications for Advertisers

For advertisers who rely on performance-driven search and shopping campaigns, AI voice-overs offer a low-cost way to add audio to video assets without hiring a voice actor. However, the opt-out default means brands with strict audio guidelines must proactively disable the feature. The June 2026 update to include Demand Gen campaigns broadens the reach to YouTube and other video placements.

Veo, on the other hand, requires more deliberate use. Advertisers who have a library of product images but limited video production resources can now generate video content algorithmically. The quality of Veo output varies; Google recommends testing multiple generations to find the best match for brand aesthetics.

Both features underscore a larger trend: Google is automating creative production to lower the barrier for video ads, especially for small and medium businesses. The trade-off is a loss of human nuance in voice delivery and visual composition. Advertisers should monitor performance data closely and consider A/B experiments before scaling AI-generated assets.

Conclusion

Google Ads' AI voice-over and Veo generative video represent two distinct approaches to reducing friction in ad creation. The voice-over feature is now a default enhancement in Performance Max and Demand Gen campaigns, with controls to opt out and measure impact. Veo offers a more experimental way to create video from images. As of August 2026, both tools are active and evolving. Advertisers should test them cautiously, using the available reporting and A/B testing features to ensure they improve campaign performance without compromising brand quality.