bug-bounty518
xss347
exploit237
google218
rce177
facebook167
microsoft138
web3122
writeup114
cve102
malware101
open-source91
csrf84
apple82
account-takeover78
sqli68
ai-agents63
browser63
cloudflare60
dos59
phishing56
ssrf52
privilege-escalation52
tool46
supply-chain45
privacy44
pentest42
reverse-engineering41
auth-bypass40
oauth38
idor38
llm37
aws36
opinion35
cloud35
ctf34
automation33
machine-learning32
race-condition32
code-generation31
infrastructure31
lfi30
clickjacking29
node28
cors28
access-control27
subdomain-takeover26
rust24
performance-optimization24
info-disclosure24
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