Short answer: No — not really.
But the picture is more complicated than that. Across the UK tech sector, a growing debate is taking place around AI scraping, AI training data ethics, and developer copyright. Large language models and AI coding tools are trained on vast amounts of publicly available code, raising serious questions about consent, licensing, and intellectual property.
Modern AI coding tools are trained on enormous datasets that include public GitHub repositories, open-source projects, coding forums and technical documentation. This process, often referred to as AI code scraping, allows models to learn patterns, syntax, and best practices.
From a technical perspective, this has produced impressive results. AI coding assistants can generate boilerplate code, debug common errors and accelerate development workflows.
From a developer’s perspective, the ethical implications of AI training data ethics are harder to ignore. Much of this code was written for collaboration, education, or open-source contribution, not for commercial AI training.
How AI Models Learn From Developer Content
Most large AI models are trained on vast quantities of publicly accessible material, including:
- Open-source repositories
- Code shared on developer forums
- Documentation and tutorials
- Q&A platforms such as Stack Overflow
This process allows models to learn syntax, logic patterns and problem-solving approaches. The result is undeniably useful tooling: faster prototyping, fewer repetitive tasks, and quicker debugging.
At the same time, it intensifies concerns around AI data scraping consent, particularly where developers were never informed their work could be used for AI training.
Why Develops Feel AI Is Stealing Their Work
The concern isn’t simply that AI tools learn from code. It is how that learning happens. Across the UK, developed consistently raise three issues when discussing developer copyright and AI.
- Lack of Consent
Most developers were never asked whether their work could be used as AI training data. The assumption that publicly accessible code is fair game has become one of the most criticised aspects of modern AI development. - Copyright and Open-Source Licensing
Open-source does not mean copyright-free. Develops worry that with AI-generated code copyright issues may arise when outputs reproduce licensed code patterns. This creates legal risk for both develops and businesses deploying AI-generated code in production. - Commercial Gain Without Compensation
AI platforms generate significant revenue, yet the developers who work contributed to training datasets typically receive no attribution or compensation. For many, this feels less like collaboration and more like extraction.
Are Developers Still Using AI Coding Tools?
Yes — and this is where the debate becomes complicated.
AI coding tools are now deeply embedded in everyday development workflows, despite ethical concerns. Many UK developers describe themselves as cautious but pragmatic. AI speeds up repetitive tasks, supports learning new frameworks and reduce time spent on routine debugging.
This creates an unresolved tension. Developers rely on tools built using practices that they don’t fully agree with.
When AI Scraping Becomes Impossible to Ignore
Recent events have brought AI intellectual property risks into sharper focus.
Reddit publicly accused Perplexity, a multi-billion-dollar AI company, of using its content despite explicit restrictions; they alleged that Perplexity accessed Reddit material indirectly by scraping Google search results, bypassing technical protections such as robots.txt.
Reddit claims it had already issued a cease-and-desist months earlier. Instead of citations declining, they reportedly increased. From Reddit’s perspective, this wasn’t accidental ingestion and instead was systematic circumvention. Cases like this intensify scrutiny around AI code scraping practices and transparency.
How This Compares to the AI Art Debate
The controversy surrounding AI-generated art mirrors the debate around AI and developer IP rights. Artists have objected to their work being scraped and used for AI training without consent, especially where outputs closely resemble original styles. In both cases, creators argue that AI systems are built upon unpaid creative labour.
There are differences:
- Visibility: AI art outputs are obvious, while code reuse is often buried inside software.
- Licensing complexity: Open-source software licenses introduce legal layers not always present in art.
- Economic impact: Developers are often salaried, though that does not remove IP concerns
At the core, both debates revolve around ethical AI development and respect for intellectual property.
Why UK Businesses Should Care
For UK organisations, this debate is not theoretical.
Using AI-generated code without understanding its origins exposes businesses to AI intellectual property risks, including:
- Copyright disputes
- Breaches of open-source licensing
- Reputational damage
- Increased regulatory scrutiny as UK AI laws evolve
As legal risks of AI-generated code become more visible, businesses need clear internal policies governing AI usage.
Best Practice: Responsible Use of AI in Development
To reduce risk while still benefitting from AI, businesses should adopt responsible use of AI in development practices:
- Be transparent about when AI tools are used
- Ensure all AI-generated code is reviewed by humans
- Understand licensing implication of AI outputs
- Prioritise tools with clear training data policies
AI should function as an assistant, not an author.
The Path Forward
The question isn’t whether AI will continue to rely on scraped data — it already does.
The real issue is how businesses, platforms and developers choose to respond as the legal and ethical frameworks evolve. For developers, discomfort is understandable. Their work powers tools that generate enormous value, often without consent, attribution, or compensation.
For businesses, the challenge is different but no less serious: adopting AI without understanding its underlying processes introduces legal, reputational, and compliance risks that can’t be ignored.
Cases like Reddit versus Perplexity, alongside other similar disputes in the creative industries, suggest we’re moving toward a world where copyright is less about absolute control and more about negotiation, transparency, and accountability. Until clearer rules exist, organisations that treat AI as a shortcut rather than a tool will be the most exposed.
The safest path forward is learning to use AI responsibly. That means understanding data provenance, respecting licensing, keeping humans in the loop and being transparent about how AI supports, rather than replaces, expertise.
Developers may not be entirely comfortable with how AI is trained today. Still, the choices businesses make now will help determine whether AI becomes a collaborative tool that still respects intellectual property or another system that quietly erodes it.