Google Photos Proprietary AI Model Extraction
Google Photos
Researchers recovered TensorFlow Lite models from the Google Photos Android application using static reverse engineering and dynamic instrumentation. The extraction provided white-box access to proprietary Google model artifacts.
- first report
- 30 Jun 2026
- last update
- 30 Jun 2026
Timeline · 2
Sources · 2
News and write-ups
- 2026-06-30Google Photos AI Models: The Secret Sauce That Can Be Stolenskyld.io · skyld.io/google-photos-model-extraction
Research and catalogues
- 2026-06-30Google Photos AI Model ExtractionMITRE ATLAS · atlas.mitre.org/studies/AML.CS0058
Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.
Classification
- lane
- Research
- AI role
- Attack on AI
- categories
- Model theft / extraction
- severity
- 1 · Research / proof of concept
- targets
- owasp llm top 10
- LLM10:2025 Unbounded Consumption
- dates
- happened Mar 2025 (approx.) · first reported 30 Jun 2026
- origin
- MITRE ATLASDerived from MITRE ATLAS (Apache 2.0, © MITRE).
- record
- confidence 0.95 · created 2026-10-18 06:55Z · updated 2026-10-18 06:55Z
Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned
Copy a brief
This record as a short Markdown brief, for a ticket, a chat or a coding agent.
the brief as text
# Google Photos Proprietary AI Model Extraction https://gcdradar.vercel.app/incidents/google-photos-proprietary-ai-model-extraction · AIR-2026-0353 · closed 2026-07-14 - Lane: Research - Severity: 1 (Research / proof of concept) - Fix status: Fix status unknown - First reported: 2026-06-30 (happened 2025-03) - AI role: Attack on AI - Products: Google Photos ## Summary Researchers recovered TensorFlow Lite models from the Google Photos Android application using static reverse engineering and dynamic instrumentation. The extraction provided white-box access to proprietary Google model artifacts. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2026-06-30: Google Photos AI Models: The Secret Sauce That Can Be Stolen (https://skyld.io/google-photos-model-extraction) - 2026-06-30: Google Photos AI Model Extraction (https://atlas.mitre.org/studies/AML.CS0058) ## Sources - 2026-06-30 MITRE ATLAS: Google Photos AI Model Extraction https://atlas.mitre.org/studies/AML.CS0058 - 2026-06-30 skyld.io: Google Photos AI Models: The Secret Sauce That Can Be Stolen https://skyld.io/google-photos-model-extraction Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closedChina-Based AI Companies Conduct Distillation Campaigns Against U.S. AI Models
Breaches & attacks · Exploited · since 2024 (approx.)
AI - liveMixture-of-Experts LLMs Characterizing Recoverable Expert-Utilization Signatures in Fine-Tuned MoE LLMs
Research · since 2 Nov 2026
AI - closedQuantized Deep Neural Networks Revisiting privacy risks and membership inference attacks in quantized DNNs
Research · since 28 Oct 2026
AI - closedOpenAI API Flaw in OpenAI Compaction Process Enables Reasoning Extraction
Research · since 26 Oct 2026 · updated 27 Oct 2026
AI - closedPrivate Evolution Membership Inference Attack on Private Evolution Synthetic Data
Research · since 23 Oct 2026
AI - closedmedical AI models Automated Screening of Privacy Risks in Medical AI
Research · since 23 Oct 2026
AI