Voice & Audio OpenAI Opens GPT-Live-1.1 to the API and Prices Silence at 2 Cents a Minute Coding Assistants Replit Pitted 32 Claude Agents Against NumPy Tuning: 18.9x Speedup While GPT-6.2 Swarms Collapsed at Scale Security GAO Covert Tests: All Four Top Chatbots Produced Viable Attack Plans in Under 12 Hours Once Guardrails Fell Meta Meta Puts Its 2030 Net-Zero Target Under Review as AI Data Centers Push Net Emissions Up 45 Percent Business Daniel Ek Is Backing Europa Fusion, a Stealth UK Startup Led by a Former DeepMind Researcher Betting on Pulsed Power Coding Assistants Stripe Tested 14 Coding Agents on 69 Payment Scenarios: 78% of Criteria Passed, but Only 29% of Runs Met Every Requirement Security A Transcription Firm Used by Police and Prosecutors Left Interview Transcripts and Body-Cam Footage on an Open S3 Bucket News UN Killer-Robot Talks End With a 2028 Deadline and No Binding Rules. About 100 States Wanted a Treaty; the US, Russia and Israel Blocked It Voice & Audio OpenAI Opens GPT-Live-1.1 to the API and Prices Silence at 2 Cents a Minute Coding Assistants Replit Pitted 32 Claude Agents Against NumPy Tuning: 18.9x Speedup While GPT-6.2 Swarms Collapsed at Scale Security GAO Covert Tests: All Four Top Chatbots Produced Viable Attack Plans in Under 12 Hours Once Guardrails Fell Meta Meta Puts Its 2030 Net-Zero Target Under Review as AI Data Centers Push Net Emissions Up 45 Percent Business Daniel Ek Is Backing Europa Fusion, a Stealth UK Startup Led by a Former DeepMind Researcher Betting on Pulsed Power Coding Assistants Stripe Tested 14 Coding Agents on 69 Payment Scenarios: 78% of Criteria Passed, but Only 29% of Runs Met Every Requirement Security A Transcription Firm Used by Police and Prosecutors Left Interview Transcripts and Body-Cam Footage on an Open S3 Bucket News UN Killer-Robot Talks End With a 2028 Deadline and No Binding Rules. About 100 States Wanted a Treaty; the US, Russia and Israel Blocked It

Google Opens SynthID to Everyone: A Public Website That Tells You Whether an Image, Video or Audio Clip Is AI-Generated

Google Opens SynthID to Everyone: A Public Website That Tells You Whether an Image, Video or Audio Clip Is AI-Generated

Google launched synthid.com, a public website where anyone can check whether an image, video or audio clip carries a SynthID watermark — the invisible signature embedded by its Nano Banana, Veo and Lyria models. The tool was previously restricted to journalists and researchers, and Google says people already make 1 million verification requests a day through the Gemini app and Chrome.

Google on Tuesday launched a public website, synthid.com, that lets anyone verify whether a piece of media — an image, a video or an audio clip — was generated using AI. The tool was first made available to selected journalists, media professionals and researchers for testing at Google I/O last year; it is now open to everybody.

The site supports a broad set of formats: JPG, JPEG, PNG, BMP, WEBP, AVIF, HEIC, HEIF, TIFF, TIF and GIF for images; MP4, MOV and WEBM for video; and WAV, MP3, OGG, FLAC, AAC and M4A for audio. Users upload a file and the service checks for the presence of a SynthID watermark, the invisible signature Google DeepMind introduced in 2023 and now embeds in media produced by its generative models.

That watermarking footprint is wide. Google's Nano Banana, Veo and Lyria models, along with tools with generation capabilities such as Gemini, Flow, ProducerAI and Vids, all watermark the media they produce. OpenAI, NVIDIA and Kakao also support SynthID, and OpenAI maintains its own site for verifying content; Apple is said to be adding support soon as well.

Verification is no longer confined to the new website. Google has baked SynthID checking into the Gemini app and Google Chrome, and said people currently make 1 million requests to verify content every day — a figure that quietly documents how much AI-generated material circulates, and how often someone downstream wants to know whether it is real.

The company is careful not to oversell the technology. Watermark-and-verify systems are not infallible: Reuters reported in July that Meta's own AI image detector failed to identify some of its own cropped AI-generated images, and SynthID's detection can degrade when media is edited, re-encoded or cropped aggressively. Microsoft and Meta maintain their own watermarking and verification standards, which means the ecosystem is fragmenting into competing provenance schemes rather than converging on one.

Still, opening the detector to the public changes the equation in one practical way: provenance checking stops being a specialist capability and becomes something an editor, a teacher or an ordinary user can do in a browser tab. As elections, stock-photo libraries and social feeds fill with synthetic media, the bottleneck has been less about watermarking than about who can actually check. Google just removed that gate — at least for the substantial share of AI content that carries its watermark.

Comments (0)

Log in to join the discussion

Log In

No comments yet