The news: While major companies are picking up generative AI (genAI) for coding, many developers remain skeptical about using it without human oversight.

• Three-quarters (78%) said AI tools have made them more productive, but a similar share (76%) don’t entirely trust AI-generated code, per Qodo’s The State of AI Code Quality report.
• One-quarter of senior developers are confident enough to ship AI code without reviewing it first.

Ramping up: Developers are integrating genAI into their workflows at a high rate, and over half (57%) said it makes their job more enjoyable.

• 82% of developers use genAI coding tools on a daily or weekly basis, and only 7% never do.
• Two-thirds (65%) said AI contributes to at least a quarter of their code.

Coders are also using multiple tools, suggesting no single company is dominating the market for genAI coding assistants. More than half (59%) use three or more AI tools regularly, and 20% use at least five.

Speed and quality: Efficiency gains from AI often go hand-in-hand with improved code quality.

Among developers who experienced significant efficiency gain, over two-thirds (70%) said that code quality got better. Even modest productivity gains led 51% to see better code.

Confidence divide: Despite efficiency and accuracy gains, hallucinations problems remain, making code reviews an integral part of the process.

• Most developers (76.4%) see high rates of hallucination and remain wary enough to not ship code without reviewing it.
• Even among users whose genAI output has low rates of hallucinations, 11.8% have low confidence.
• Only 3.8% experience the ideal situation: Few hallucinations and high trust in AI’s output.

Yes, but: Coders who are adopting genAI may be training it to take over their jobs. Layoffs and threats to job security may be accelerated by companies’ interest in freeing up resources for AI spending.

• Half of AI experts and 48% of US adults believe that AI will lead to fewer software engineering jobs over the next 20 years, per Pew Research.
• Over 63,000 tech jobs have been cut so far this year, per Layoffs.fyi.

Our take: Remaining skepticism from developers—one of the professions closest to AI—shows that companies use genAI as a support tool and co-pilot rather than a replacement for human judgment. Training employees on AI’s weaknesses and requiring review can help reduce errors.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.