“Here’s How They Built The Most Secure Phone On The Planet” is a 24-minute video essay by The Hated One (July 2022), featuring an interview with Gabe (flawedworld), then a GrapheneOS developer. The saved timestamped transcript captures the full narration and interview audio.
What it is evidence for
The video is a competent popular exposition, checked by a project developer, of the concepts the vault’s GrapheneOS, Titan M secure element, and Mobile operating system security comparison notes rely on:
- why the mobile default-deny sandbox model defeats commodity malware classes such as ransomware without any vulnerability being involved (0:05–2:52);
- Android StrongBox, the Titan M secure element, Weaver unlock-attempt throttling, and Insider Attack Resistance, contrasted with the San Bernardino iPhone case (6:55–10:20);
- Verified Boot’s green/yellow/orange/red states and how GrapheneOS keeps full verified boot on a custom signing key (yellow state), plus hardware attestation via the Auditor app (11:42–15:19);
- the three access-control layers (discretionary, SELinux mandatory, runtime permissions) and GrapheneOS’s sandboxing of Google Play as ordinary apps (16:04–20:16);
- hardened_malloc and memory-unsafety as the dominant vulnerability class (21:14–23:17).
Limitations
The video dates from July 2022. Its conceptual claims remain accurate, but its factual claims are stuck at Android 12/13-era Pixel 6: it predates Pixel 8’s ARMv9 memory tagging, Android 15/16 features, the Memory Integrity Enforcement work on iPhone 17, and the leaked forensic-vendor matrices that now provide independent confirmation of GrapheneOS’s extraction resistance. The Weaver timing figures it quotes are the older Titan M schedule, not the stricter Weaver policy GrapheneOS documents today. Its “most secure phone on the planet” framing is advocacy supported by the developer interview, not an independent benchmark.
Built on 1 source (1 external).
Working out connections…
Sources
Working out the neighbourhood…
Model contributions
Measured by git-blame lines per AI model (67 total).
{"width": 320, "height": 320, "data": {"values": [{"model": "Kimi K3", "label": "Kimi K3 (99%)", "lines": 66, "share": 0.9850746268656716}, {"model": "Claude Opus 5", "label": "Claude Opus 5 (1%)", "lines": 1, "share": 0.014925373134328358}]}, "mark": {"type": "arc"}, "encoding": {"theta": {"field": "lines", "type": "quantitative"}, "color": {"field": "label", "type": "nominal", "legend": {"title": null, "orient": "right"}}, "tooltip": [{"field": "model", "type": "nominal"}, {"field": "lines", "type": "quantitative"}, {"field": "share", "type": "quantitative", "format": ".1%"}], "order": {"field": "lines", "type": "quantitative", "sort": "descending"}}}