Can anyone create a personal avatar in Google Vids?

No. According to TechCrunch, personal avatar access is limited to users in certain regions who are aged 18 or older, and avatars are tied to the account holder's likeness and Google account.

What happened to OpenAI's Sora?

Sora was shut down, though TechCrunch does not provide details on the shutdown timeline or reason in this article.

How does Gemini Omni work in Vids?

Gemini Omni accepts written prompts and reference images, then combines those inputs to generate AI videos. It can also edit existing phone videos by adjusting backgrounds, lighting, or adding effects.

Are personal avatars watermarked?

Yes. According to TechCrunch, avatars are watermarked invisibly with SynthID to prevent misuse.

Google Vids Adds Personal AI Avatars and Gemini Omni Integration

Custom Avatars and Expanded Capabilities

Google announced on July 16 a suite of updates to Google Vids that moves the platform from a specialized workplace presentation tool into a broader content-creation ecosystem. According to TechCrunch, the company is introducing custom digital avatars that users can generate from a selfie and voice recording, allowing individuals to appear as on-camera talents in their own videos without physical recording.

The rollout also includes Gemini Omni integration, which enables multi-modal video generation by combining written prompts with uploaded reference images. According to TechCrunch, this model can also edit existing footage—adjusting backgrounds, correcting lighting, or layering effects—without requiring users to re-record from scratch.

Competitive Positioning Against Specialized Platforms

The expansion signals Google’s intent to challenge dedicated AI video startups. According to TechCrunch, Google’s feature set now positions Vids in direct competition with platforms like HeyGen, Synthesia, Captions, and D-ID, which have built narrower but deeper toolsets around avatar generation and video synthesis. By embedding these capabilities into Google Workspace, Google is leveraging its distribution advantage to make AI video creation a standard office function rather than a specialized procurement decision.

Safety Controls and Access Restrictions

Google has implemented guardrails around avatar misuse. According to TechCrunch, personal avatars are account-tied, watermarked with SynthID (an invisible identifier to track AI-generated content), and restricted to users aged 18 or older in select regions. TechCrunch notes this prevents the kinds of unauthorized deepfake scenarios that occurred with OpenAI’s Sora, which previously allowed users to generate realistic impersonations of public figures including OpenAI CEO Sam Altman.

Why This Matters

For enterprise video producers: Vids’ step-by-step editing workflow reduces friction for non-technical teams creating training videos, product demos, or internal communications. The personal avatar feature removes the need for on-camera talent or external production, lowering cost per video while maintaining personalization at scale.

For the AI video market: Google’s entry into the platform layer threatens the premium positioning of specialized startups. HeyGen and Synthesia have built investor-backed businesses on avatar + editing workflows; if Google’s implementation reaches feature parity and distributes through Workspace’s installed base (estimated at over 100 million paying seats), pricing pressure and consolidation in the segment becomes likely.

For content authenticity: The SynthID watermarking raises an open question about detection robustness. If an avatar video is edited, stripped, or re-encoded by a third party, SynthID’s invisibility only protects Google’s interests if detection systems work reliably at scale—a claim that remains unvalidated outside lab conditions.