REVIEWS / AI CODING / OWNER INSIGHTS
🦉 WE READ 129 OWNER COMMENTS
GitHub Copilot Business: what owners actually say
Owners appreciate Copilot as a documentation shortcut and 'second brain' for familiar stacks, but question its correctness, its licensing hygiene, and whether Microsoft's real play is habit-formation before lock-in.
What owners complain about
- Opt-out data training default COMMON
Users are automatically opted into allowing GitHub to use their data for AI model training; the toggle is buried under settings and must be manually disabled.
- Confidently wrong suggestions SOME
At least one developer working in C# with Visual Studio reports Copilot frequently producing incorrect code that took longer to fix than writing from scratch, calling it 'completely useless' for their non-esoteric work.
- AI slop feedback loop SOME
Developers worry repos already full of AI-generated code will train future models on that same low-quality output, compounding slop.
- GPL licensing concerns FEW
Users ask whether Copilot still reproduces GPL-licensed code, reflecting ongoing unease about legal compliance for business use.
- Ceiling on generalization FEW
One commenter notes Copilot-like models produce output similar to training inputs, with a ceiling on how well they can generalize—making them unsuitable for genuinely novel tasks.
What owners love
- Second-brain codebase awareness
Developers report Copilot learns their codebase and coding style over time, so a simple comment like 'upsert the object in the store' produces usable multi-step logic in JS/Vue and Python.
- Documentation cache-miss eliminator
For developers working frequently with APIs and high-level code, Copilot saves time by surfacing method signatures and input formats so they don't have to look up documentation repeatedly.
- Free Claude 3.5 Sonnet access
The inclusion of 50 monthly chats with Claude 3.5 Sonnet—a direct competitor's model widely regarded as among the best for coding—is called 'highly surprising' and a genuinely valuable perk.
Surprising patterns
- The single most-upvoted sentiment across all platforms is a drug-dealer metaphor ('First hit's free'), signaling that developers see the business model—cheap or free individual access leading to enterprise upsell via internal advocacy—as transparent and somewhat manipulative.
- Several commenters explicitly describe a strategy where Microsoft wants individual devs to form AI habits and then pressure their managers to buy licenses, which users view as deliberate end-run around stalled enterprise adoption.
- Ideological opposition to AI in general (ethically bankrupt, 'killing people,' enshittification) is concentrated on Lemmy and often comes from non-users, whereas Hacker News and Reddit users who actually use the product focus their criticism on concrete shortcomings like incorrect output and licensing risk.
WHO SHOULD SKIP IT
Developers working on novel or non-standard codebases who need reliably correct output—commenters report Copilot's suggestions can be confidently wrong and take more time to fix than writing the code yourself.
Synthesised from 129 real owner comments across 6 platforms. Every point is grounded in the comments — no marketing, no AI guessing. How we do it →