AI exposes decades of human coding flaws, sparking repair push

The "Black October" cybersecurity crisis stems from decades of neglected human-written software, not AI-generated code, exposing vulnerabilities that AI tools are now rapidly identifying. This has sparked industry debate about who is responsible for fixing the internet's fragile foundations. AI's ability to scan software at speeds humans cannot match has revealed security weaknesses that were long overlooked or tolerated, with companies often justifying never fixing issues. The surge in published vulnerabilities reflects AI exposing existing human slop rather than creating new problems, as volunteers and small groups maintain critical infrastructure. Tech leaders like Vercel's Tom Occhino propose a coalition to fund AI-driven repairs of vulnerable software, similar to OpenAI's Patch the Planet effort. This shift toward fixing existing systems, rather than chasing new ideas, offers a constructive response to the crisis.