No More Ugly Watermarks on Your AI Images — But Credit Still Applies

Have you ever asked an AI tool to create a promotional image for your new product, only to find a small “AI” stamp stuck in the corner? If you run a Malaysian SME, you’ve probably spent more time than you’d like cropping, covering, or accepting that imperfect mark on your social media posts, brochures, or product listings. It makes otherwise professional-looking content feel second-hand.

Google just changed that. On August 14, the company announced that users can now remove the visible watermark from AI-generated images, videos, and songs. The toggle will be available in Gemini and Flow (Google’s video editor), with Search support coming soon. For business owners who rely on quick, low-cost content creation, this sounds like a gift. But before you celebrate, there’s something important to understand: the watermark may disappear from your eyes, but it doesn’t disappear from the file itself.

TL;DR: Google now lets you turn off the visible watermark on AI-generated images, videos, and songs from Gemini and Flow. Invisible SynthID watermarks and C2PA metadata still stay embedded. You get cleaner content — but you also inherit the responsibility of knowing exactly what’s AI-made.

What This Means

Let’s break this down without the jargon. A visible watermark is the small logo or label you see on an image, video, or audio track. It tells people “this was made by AI.” An invisible watermark is a digital fingerprint embedded in the file’s data. You can’t see it, but software can. C2PA metadata is like a nutrition label for media: it records the origin, the tools used, and any edits made to the file.

When you remove the visible watermark in Google’s tools, the invisible SynthID watermark and C2PA metadata remain untouched. That means anyone with access to Google’s detection tools — like Search or a SynthID-enabled system — can still determine whether content was AI-generated. In fact, Google’s VP for Gemini, Josh Woodward, said it directly: “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.” (TechCrunch)

“Removing the visible watermark is like wiping the sweat off a product label — the invisible barcode is still there, and the system still knows where it came from.”

How This Applies to Malaysian SMEs

For Malaysian SMEs, this update isn’t just about aesthetics — it opens real commercial opportunities. Consider the e-commerce seller on Shopee or Lazada. Product photos are the main driver of clicks and sales. A visible watermark on a hero image makes your listing look generic and untrustworthy. With the new Google setting, you can generate clean, professional product shots and lifestyle images without that awkward “AI” badge. The result is a more polished storefront that matches the quality of larger brands — and you can produce it in minutes, not days.

Think about the F&B business — hawker stalls, cafes, or catering companies. You want to post daily specials or promotions on Instagram and TikTok. Food images with visible watermarks often look like stock photos or, worse, fake AI food. By removing the mark, your content feels more authentic, which matters for local customers who care about genuine reviews and photos. That same logic applies to travel agencies, homestay operators, and event planners: every visual asset you share represents your brand. A clean image builds confidence; a watermarked one creates doubt.

Then there’s video. Flow, Google’s video editor, now allows watermark-free output for videos made with the Omni or Lyria models. Malaysian SMEs can use this to create product demos, training videos, or location walkthroughs for their staff and customers. For example, a fitness centre could generate a short workout guide video without a distracting logo, making it more engaging for their community. Or a tuition centre could create animated lesson previews for their website. The creative possibilities are genuinely useful for small teams that don’t have a video editor on payroll.

But there’s also a cautionary side. Just because you can remove the visible mark doesn’t mean you should hide the fact that you used AI. Malaysian customers value honesty. If you pass off AI-generated food photos as real dishes, you could lose their trust when the actual food looks different. The invisible metadata still keeps you accountable. This is not about fooling people — it’s about making AI content look better while being transparent. Use the watermarked-free output for polished work, but be ready to disclose AI use if a customer asks. That’s the ethical line.

Type of Asset	With Visible Watermark (Before)	Without Visible Watermark (After)	Recommended Practice for SMEs
E-commerce product images	Looks like a draft; lower conversion	Clean, professional look	Use for listings; keep your own colour or style consistency
Social media ads (images)	Distracts attention from your offer	Better integration with ad creative	Pair with clear text overlays and brand colours
Video content for Facebook/YouTube	Visible mark reduces watch time	More immersive experience	Still add your own logo or end-screen
Background music (Lyria)	Audio “watermark” may be awkward	Cleaner audio for branding	Keep a record of the source for licensing clarity

Practical Takeaways

• Go to Settings > Media Watermark in Gemini or Flow to turn the visible mark on or off. It’s a per-project setting, so decide based on the purpose of the content.
• For public-facing marketing assets, remove the visible watermark to improve brand perception. But always keep the invisible metadata intact — it protects you if someone accuses you of hiding AI usage.
• If you receive AI-generated files from freelancers or suppliers, use Google’s Search or SynthID detection tools to verify if the content is AI-made, especially for sensitive claims like product reviews or before-after photos.
• Set a simple policy within your team: “We use AI for support materials, but we disclose any AI-generated content that influences customer decisions.” It’s easier to stay honest when the policy is in writing.
• Do not rely solely on the visible watermark to prove authenticity. Since it can be removed, your customers may not know the content is AI — so you have to be the one to tell them if it matters.

The Bigger Picture

Google’s move is a clear shift in how tech companies approach AI transparency. Instead of forcing a visible mark on everything, they’re moving toward a “hidden but traceable” model. This is more practical for businesses, but it also places more responsibility on content creators. In the long run, you’ll see more AI-generated content in marketplaces, in ads, and in customer communications. The visible label will become rarer, but the invisible tools to detect it will become more common.

For Malaysian SMEs, this means two things. First, your content quality bar just got higher. Competitors can now produce clean AI images and videos too, so you need to focus on your unique product, story, and customer experience to stand out — not just a clean visual. Second, your digital responsibility is growing. With watermarking becoming optional, the trust between you and your customers will increasingly rely on your honesty rather than on a small icon in the corner.

This is not a disaster or a warning. It’s simply the natural evolution of AI tools in business. Use the new capability to make your business look better, but use it with integrity. The patients will appreciate the clearer labels, and the savvier customers will appreciate your transparency.

Ready to Streamline Your Operations?

Your business should run itself. AutoRunBiz deploys AI agents to automate your daily operations — WhatsApp orders, invoicing, customer follow-ups, and accounting. Book a free 15-min ops audit to see where automation fits your business →