On AI-powered hackers: AI so far fixes vulnerabilities more than it creates break-ins
binarybits · x · 2026-11-18
Responding to fears of AI-armed hackers, Michael Benedikt offers a marker: don't confuse people locking open doors with robbers breaking in. AI-equipped tools may make hackers more dangerous, but AI's record so far is solving technical debt and covering existing vulnerabilities — e.g., the PHP patch was nearly done before humans blabbed and gave black hats an opening.
More from Safety
- Capping RSI at 10% of Compute: A Proposal to Mandate 90% for Public Benefit — trevposts · 2026-11-18
- If Firms Can't Stop China Stealing Weights, They Shouldn't Build Risky AI — trevposts · 2026-11-18
- If Models Know When They're Being Evaluated, Can Alignment Evals Still Be Trusted? — trevposts · 2026-11-18
- LLMs Let Attackers Diff Patches Into Exploits in Minutes—Patch Now — danielrock · 2026-11-18
- Microsoft launches AI transparency initiative with 40+ orgs, Foundry Hosted Agents coming — testingcatalog · 2026-11-18
- US officials: Megaspeed probed since 2024, Nvidia put it on no-sell list in June 2025 — Eliv_nurotic · 2026-11-18