GitHub Weighs Pull-Request Controls as Engineers Warn 75% AI-Written Code Can Turn to Slop
Updated
Updated · InfoWorld · May 25
GitHub Weighs Pull-Request Controls as Engineers Warn 75% AI-Written Code Can Turn to Slop
3 articles · Updated · InfoWorld · May 25
Summary
GitHub is considering tighter pull-request filters and maintainer controls after open-source projects were swamped by low-quality, often AI-generated contributions.
Mario Zechner and Armin Ronacher, engineers behind parts of OpenClaw, say AI coding tools produce “vibe slop” by letting users skip design, testing, judgment and ownership while generating plausible but flawed code at scale.
The risk is less obviously broken output than “80% right” code that slips into production, where review capacity does not scale with fleets of coding agents churning out changes.
The report argues AI works best on bounded tasks such as tests or code explanation, while broad architectural decisions still require experienced engineers and enforced guardrails.
Google says 75% of its new code is AI-generated and engineer-reviewed, underscoring the broader debate over whether AI boosts disciplined teams or accelerates weak engineering practices.
As AI floods open source with 'vibe slop,' is it destroying the very communities it was meant to empower?
With AI writing most new code, are we training engineers to be expert critics or just glorified rubber-stampers?
"Vibe Slop Crisis: Over 30% of AI-Generated Code Contains Critical Vulnerabilities—Security, Economic, and Regulatory Impacts in 2026"
Overview
The report highlights an urgent crisis in the digital world called "Vibe Slop," where AI-generated code is created rapidly without enough security checks or human review. This has led to a sharp rise in software vulnerabilities and maintenance problems, putting heavy pressure on development teams. A key example is the Moltbook breach in early 2026, where a social network built quickly with AI suffered a major data leak due to poor security controls. As more platforms face similar risks, the industry is scrambling to adopt stronger defenses and rethink how AI-driven development should be managed to protect users and maintain trust.