bug-bounty521
xss350
exploit239
google226
rce179
facebook170
microsoft144
web3122
writeup114
malware109
cve106
open-source91
csrf85
apple83
account-takeover80
sqli68
ai-agents63
browser63
cloudflare60
dos60
phishing58
ssrf52
privilege-escalation52
tool46
supply-chain46
reverse-engineering46
pentest46
privacy44
auth-bypass41
idor38
oauth38
llm37
aws37
ctf36
cloud36
opinion35
race-condition33
automation33
lfi32
machine-learning32
infrastructure31
code-generation31
node29
clickjacking29
cors28
access-control27
subdomain-takeover26
react25
performance-optimization24
rust24
0
2/10
A mathematical modeling study examining how AI-assisted code generation amplifies technical debt without proportional QA investment, using coupled ODEs to define a critical QA threshold (η_c = 4γv) below which systems collapse into unrecoverable technical debt. Empirical calibration across 1.5M file-touch events shows AI code erodes validation capacity 12× faster than human code, but a single dedicated tester yields 18:1 ROI.
ai-assisted-development
code-quality
technical-debt
software-engineering
dynamical-systems
qa-methodology
empirical-research
git-analysis
Antonio Mennillo
DOI 10.5281/zenodo.18971198