#USA #San Francisco #AI Software Delivery #DevOps Maturity #Code Production

The Impact of AI on Development Speed and DevOps Maturity: New Insights from Harness Study

On March 11, 2026, Harness, the AI Software Delivery Platform™, unveiled significant findings regarding the influence of artificial intelligence (AI) on software development processes, revealing both advantages and pitfalls. Their study indicates that while AI technology is spurring faster code creation, the essential delivery systems that ensure code quality and security are lagging behind.

This comprehensive report, based on feedback from 700 engineers and technical managers across several countries, including the U.S., U.K., France, Germany, and India, suggests that increased deployment speeds are contributing to systemic issues within DevOps environments. These challenges include elevated risks during deployment, a rise in manual workloads post-code writing, heightened burnout rates among engineers, and longer times required to resolve incidents.

Key Findings from the Harness Report

The report presents some startling statistics:

• - Faster Deployments: 45% of developers utilizing AI coding tools multiple times daily are deploying code to production daily or even faster. In contrast, only 32% of daily tool users and 15% of weekly coding tool users can achieve this speed.
• - Deployment Problems: A whopping 69% of frequent AI coding tool users reported experiencing ongoing deployment issues, with 58% of all respondents echoing these concerns.
• - Incident Recovery: Teams engaged with AI coding tools on a daily basis typically take an average of 7.6 hours to address production incidents, compared to 6.3 hours for less frequent users.
• - Increased Manual Work: 47% of frequent AI coding users indicated that they encounter more manual tasks such as quality assurance, remediation, and validation post-code creation.
• - After-Hours Work: An alarming 96% of heavy users reported that they often had to work evenings or weekends because of release-related duties, compared to 66% of occasional users.

The findings reveal a clear trend: while AI expedites code generation, it simultaneously places increased pressure on development teams. Individuals spend a staggering 36% of their time on tedious manual tasks that can include configuration, approvals, and resolving failures, which only exacerbates workplace stress and burnout.

The Need for Modernization and Standardization

Trevor Stuart, SVP and General Manager at Harness, noted that while AI coding tools accelerate the development tempo, the overall delivery pipeline remains under-resourced. His observations suggest that organizations are eager to leverage AI for generating code but are often neglecting to modernize critical delivery processes and automation practices that align with this new pace of work.

The study revealed several concerning trends regarding development teams’ methodologies:

• - 73% of engineering leaders admitted their teams lack standardized templates or guides for services and pipelines, leaving them vulnerable to inconsistencies.
• - Only 21% could successfully implement functional build and deployment pipelines within a two-hour timeframe.
• - An overwhelming 77% reported needing to defer code shipping due to reliance on other teams for routine delivery processes.

Furthermore, the report highlights that with accelerated deployment speeds, inadequate quality and security controls may lead to increased vulnerabilities. These risks accentuate the necessity of a more robust foundation upon which AI-driven strategies can operate efficiently.

Recommendations for Optimization

As organizations strive to capitalize on AI's capabilities, it's essential to prioritize the modernization of development frameworks. Stuart suggests three focal points:

1. Standardize Delivery Pipelines: Establishing consistent templates and 'golden paths' can facilitate safer and more reliable application deployments.
2. Automate Quality and Security Checks: Implementing controls earlier in the software development lifecycle can help catch potential problems before they escalate to production.
3. Implement Safety Guardrails: Utilizing features like automated rollbacks and centralized secrets management can mitigate impacts stemming from unforeseen failures.

Through these measures, companies can not only enjoy the speed benefits associated with AI-driven development but do so while maintaining integrity, security, and the overall well-being of engineering personnel.

Conclusion

With the landscape of software delivery evolving rapidly under the influence of AI, it is paramount for enterprises to adapt their development practices accordingly. The findings from the Harness study serve as a crucial reminder that speed should not come at the expense of quality or team well-being. Embracing a holistic approach to the software delivery lifecycle, ensured through modernization and standardization, will be the key to thriving in this new era of AI-enhanced development.