Google Made AI Watermarks Optional — Should Your Business Care?
Imagine spending an hour generating the perfect product photo for your online store — only to realise the AI tool stamped a faint “Gemini” badge across the corner. For many Malaysian business owners, that watermark was the difference between a usable asset and one that made your brand look amateurish. That friction just disappeared.
On August 14, 2026, Google announced it will now allow users to remove the visible watermark from its AI-generated images, videos, and songs. If you run a small business in Malaysia — whether you’re selling handcrafted batik on Lazada, running a boutique design studio in Penang, or managing social media for a furniture brand in Johor — this change reshapes how you can use AI in your daily operations.
What Happened
Google’s vice president for Gemini, Josh Woodward, posted on X that the toggle to turn off visible watermarks will be available for the Nano Banana, Omni, and Lyria models. The setting appears in Gemini and Google’s video editor, Flow, with Search support coming soon.
Here’s the critical nuance: this removes only the visible watermark. The invisible SynthID watermark and C2PA metadata remain untouched. Anyone can still verify whether content was AI-generated — the marker just no longer shouts at your customers.
“We’re striking a balance here between creative control and safety: while the visible watermarks are now optional, invisible SynthID watermarks and C2PA metadata are still being used for transparency. So you can still use Gemini or Search to see if an image was AI-generated,” Woodward wrote.
The feature rolls out in the coming days. Once available, you navigate to Settings > Media Watermark to switch the visible mark on or off. Google is also open-sourcing a new library called Credentio, allowing developers to embed a local validation mechanism in their own apps.
Why This Matters for Malaysian SMEs
Think about how much of your marketing collateral now involves AI. Product images for your e-commerce listings. Video ads for TikTok and Instagram Reels. Background music for brand videos. Before this announcement, every one of those assets carried a visual stamp that whispered to your market: “This was made by a machine, not crafted by your brand.” That subtle signal undercut the polish of your content.
Now that you can switch the watermark off, the field levels. A small team in Shah Alam can produce visual assets with the same professional finish as a corporate giant with an in-house creative department — and nothing visibly separates the two. For Malaysian SMEs, where marketing resources are stretched but customers still expect professional-grade content, this removes a real obstacle between you and a consistent brand image.
Yet the quieter story is the return to invisible, metadata-based authentication. Your customers are increasingly aware of deepfakes and AI-generated scams — from fake celebrity endorsements to fabricated news images circulating on WhatsApp. By keeping SynthID and C2PA metadata intact, Google preserves a way to verify authenticity without forcing you to label everything you publish. The practical takeaway: you gain creative freedom without losing the ability to prove your content is legitimate when it matters.
The Bigger Picture
This move also sharpens the divide in how AI companies handle transparency. It follows Anthropic’s contested decision to embed watermarks in text and files generated by Claude in response to EU regulations. Two philosophies are emerging: one relies on invisible technical provenance, the other on visible labels by default.
For Malaysian SMEs, the larger lesson concerns differentiation. When watermarks disappear, output quality becomes the only thing separating your brand from a competitor using the same tools. The businesses that win are those treating AI as an accelerator for their own taste, judgment, and customer understanding — not as a replacement for them.
Three quick actions for you this week:
- Check Settings > Media Watermark in Gemini and Flow — decide whether a visible mark serves your brand (for most commercial use, you’ll want it off).
- Review how you verify content authenticity today — if you share AI-generated images publicly, keep records of their provenance in case a customer or regulator asks.
- Watch for Credentio — Google’s open-sourced validation library could become a useful building block for Malaysian developers building AI-powered tools.
| What changes | What stays the same |
|---|---|
| Visible watermark can be toggled off in Gemini and Flow | Invisible SynthID watermark remains embedded |
| Available for Nano Banana, Omni, and Lyria models | C2PA metadata continues to track content provenance |
| Google open-sources Credentio for app developers | Search can still verify whether content is AI-generated |
Ready to Streamline Your Operations?
Technology moves fast. Your operations should keep up. AutoRunBiz builds AI systems that run your daily workflows — from WhatsApp order capture to accounting. Book a free 15-min ops audit →