--- title: Google DeepMind’s watermarked AI proteins still work in... description: Google DeepMind has introduced SynthID Bio, a watermarking technique for AI-designed proteins, embedding a detectable signature in either the amino acid... canonical: https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Google DeepMind’s watermarked AI proteins still work in the lab | daily.dev og:description: Google DeepMind has introduced SynthID Bio, a watermarking technique for AI-designed proteins, embedding a detectable signature in either the amino acid... og:url: https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9 og:image: https://api.daily.dev/og/posts/taxODj1j9.png og:image:alt: Google DeepMind’s watermarked AI proteins still work in the lab og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Google DeepMind’s watermarked AI proteins still work in the lab **[The Next Web](https://daily.dev/sources/tnw)** · 4 min read · 0 upvotes · 0 comments ## Summary Google DeepMind has introduced SynthID Bio, a watermarking technique for AI-designed proteins, embedding a detectable signature in either the amino acid sequence or the predicted 3D structure without affecting protein function. Announced in a blog post and a Nature paper, the system works inside ProteinMPNN (used with AlphaProteo) and fine-tunes part of AlphaFold 3's diffusion network. Lab tests on three binder targets (VEGF-A, SARS-CoV-2 spike RBD, PD-L1) showed watermarked designs matched unmarked ones in hit rate, binding affinity, and diversity. The tool aims to help DNA synthesis companies screen orders and label AI-made entries in databases like the Protein Data Bank. Known limitations include vulnerability to key leaks, signal dilution when fusing proteins, and incompatibility with non-ProteinMPNN tools. DeepMind has also extended the approach to Evo 2, watermarking a bacteriophage genome with Stanford's Hie lab and the Arc Institute, and is open-sourcing code and lab data. A separate SynthID dispute involves a Nikon photography contest video accused of undisclosed AI generation. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is SynthID Bio and how does it watermark AI-designed proteins? SynthID Bio is a Google DeepMind tool that embeds a detectable signature into AI-designed proteins, either by nudging amino acid choices in a sequence or adjusting atomic coordinates in a predicted 3D structure. It works inside ProteinMPNN using a cryptographic-like key to bias suggested amino acids, while ProteinMPNN still rejects any suggestion that would break the protein's function. _Teams adopting AI protein design tools can track how watermarking and provenance tracking evolve via daily.dev._ ### Does watermarking AI-designed proteins with SynthID Bio affect how they function in the lab? No, lab tests found watermarked binder proteins matched unmarked ones on hit rate, binding affinity, and sequence diversity. The tests covered three targets: VEGF-A, the SARS-CoV-2 spike protein RBD, and PD-L1, with designs generated using AlphaProteo and a SynthID Bio version of ProteinMPNN, and lab validation performed with Adaptyv Bio. _Researchers evaluating AI-generated biological designs can follow validation results like these on daily.dev._ ### What are the known limitations of DeepMind's SynthID Bio protein watermarking system? The system is only as secure as the process used to share and store its keys, and very short proteins may carry too few marked amino acids to detect reliably. Fusing a marked protein with an unmarked one can dilute the signal, detection is statistical so false positive and negative rates depend on the chosen cut-off, and many AI protein design tools besides ProteinMPNN are not yet supported. _Anyone weighing the reliability of emerging AI safety tooling can track these caveats on daily.dev._ ## Similar posts on daily.dev - [Text AI watermarks will always be trivial to remove](https://daily.dev/posts/text-ai-watermarks-will-always-be-trivial-to-remove-borbz7x9u) · sean goedecke · 6 upvotes · 4 comments - [A biological 0-day? Threat-screening tools may miss AI-designed proteins.](https://daily.dev/posts/a-biological-0-day-threat-screening-tools-may-miss-ai-designed-proteins--nr9ydrhu0) · Ars Technica · 0 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai) [View this post on daily.dev](https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Google DeepMind’s watermarked AI proteins still work in the lab","url":"https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9"},"datePublished":"2026-10-01T14:29:42.562Z","dateModified":"2026-10-01T14:30:27.629Z","description":"Google DeepMind has introduced SynthID Bio, a watermarking technique for AI-designed proteins, embedding a detectable signature in either the amino acid...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/87fbe51aed445b73e2bacc0ab60bd487?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/87fbe51aed445b73e2bacc0ab60bd487?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"The Next Web","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"The Next Web","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tnw","url":"https://daily.dev/sources/tnw"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"The Next Web","item":"https://daily.dev/sources/tnw"},{"@type":"ListItem","position":3,"name":"Google DeepMind’s watermarked AI proteins still work in the lab"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/google-deepmind-s-watermarked-ai-proteins-still-work-in-the-lab-taxodj1j9#faq","mainEntity":[{"@type":"Question","name":"What is SynthID Bio and how does it watermark AI-designed proteins?","acceptedAnswer":{"@type":"Answer","text":"SynthID Bio is a Google DeepMind tool that embeds a detectable signature into AI-designed proteins, either by nudging amino acid choices in a sequence or adjusting atomic coordinates in a predicted 3D structure. It works inside ProteinMPNN using a cryptographic-like key to bias suggested amino acids, while ProteinMPNN still rejects any suggestion that would break the protein's function. Teams adopting AI protein design tools can track how watermarking and provenance tracking evolve via daily.dev."}},{"@type":"Question","name":"Does watermarking AI-designed proteins with SynthID Bio affect how they function in the lab?","acceptedAnswer":{"@type":"Answer","text":"No, lab tests found watermarked binder proteins matched unmarked ones on hit rate, binding affinity, and sequence diversity. The tests covered three targets: VEGF-A, the SARS-CoV-2 spike protein RBD, and PD-L1, with designs generated using AlphaProteo and a SynthID Bio version of ProteinMPNN, and lab validation performed with Adaptyv Bio. Researchers evaluating AI-generated biological designs can follow validation results like these on daily.dev."}},{"@type":"Question","name":"What are the known limitations of DeepMind's SynthID Bio protein watermarking system?","acceptedAnswer":{"@type":"Answer","text":"The system is only as secure as the process used to share and store its keys, and very short proteins may carry too few marked amino acids to detect reliably. Fusing a marked protein with an unmarked one can dilute the signal, detection is statistical so false positive and negative rates depend on the chosen cut-off, and many AI protein design tools besides ProteinMPNN are not yet supported. Anyone weighing the reliability of emerging AI safety tooling can track these caveats on daily.dev."}}]} ```