cd /news/ai-safety/playstation-5-firmware-ps5-firmware-… · home › topics › ai-safety › article
[ARTICLE · art-158113] src=reversedwithai.com ↗ pub= topic=ai-safety verified=true sentiment=↓ negative

PlayStation 5 firmware: PS5 firmware security analysis using Codex and ChatGPT

Security researcher Devon used OpenAI's Codex and ChatGPT to analyze FreeBSD 11.0.0 kernel and WebKit source code in a PlayStation 5 firmware security study, answering over 130 questions with source citations, according to an open-source GitHub project. Devon reported that OpenAI's Daybreak content filter repeatedly triggered on prompts containing structural terminology and that the account was subsequently deactivated, though Devon maintains the work was strictly academic and involved no proprietary system access or exploit creation. Other researchers noted that LLM passes over JavaScriptCore can identify missing checks but the false positive rate for lifetime bugs remains high.

by read1 min views1 publishedOct 26, 2026
PlayStation 5 firmware: PS5 firmware security analysis using Codex and ChatGPT
Image: Reversedwithai (auto-discovered)

Evidence at a glance #

source evidence…

What was done #

Researcher Devon utilized Codex and ChatGPT to analyze FreeBSD and WebKit source code for PS5 security research, resulting in account deactivation due to alleged content filter triggers.

The author, Devon, conducted academic source-code analysis on the FreeBSD 11.0.0 kernel and the WebKit browser engine to investigate the security architecture of PlayStation 5 firmware. The research, documented in an open-source project on GitHub, involved using Codex and ChatGPT to trace function call chains, data structure layouts, and allocation patterns. The author reports that over 130 questions were answered by the model with source citations. The author notes that OpenAI's Daybreak content filter repeatedly triggered on prompts involving structural terminology, which the author suggests may have contributed to the subsequent deactivation of their account. The author maintains that the work was strictly academic and did not involve accessing proprietary systems or creating exploits. Other researchers have noted that while LLM passes over JavaScriptCore (JSC) can identify missing checks, the false positive rate for lifetime bugs remains high.

── more in #ai-safety 4 stories · sorted by recency
── more on @openai 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/playstation-5-firmwa…] indexed:0 read:1min 2026-10-26 · —