AI Coding, Web Coding And Agent Crowd AI Code Tools has made a dramatic and surprising recent market entry, with market value 8 4.8 billion and is expected to grow at an annual rate of 23%. Enterprises are stuck with AI coding agents and what about expensive human coders.
They are not short of advice. Openei’s CEO estimates that AI can perform More than 50% of human engineers. Six months ago, Anthropic’s CEO said that A.I 90% will write the code Six months in, Meta’s CEO said he believed Aiwal was a testament Replace mid-level engineers “soon”. To decide Recent Tech Roofsmany executives seem to be heeding that advice.
Software engineers and data scientists are among the most expensive pay lines at many companies, and business and technology leaders may be tempted to replace them with AI. However, recent high-profile failures show that engineers and their skills are valuable, even as AI continues to make impressive progress.
Saster disaster
Jason Lemkin, a tech entrepreneur and founder of SaaS Community Suster, is coding a SaaS networking app and live-tweeting his experience. About a week into his adventure, he admitted to his audience that something was going terribly wrong. Ai Deleted your production database Despite his request for a “code and action freeze”. This is the kind of mistake that an experienced (or semi-experienced) engineer will not make.
If you’ve ever worked in a professional coding environment, you know how to partition your development environment from production. Junior engineers are given full access to the development environment (this is important for productivity), but production access is given to highly trusted senior engineers on a limited as-needed basis. The reason for limited access is specifically for this use case: to prevent a junior engineer from accidentally cutting production.
In fact, Lemkin made two mistakes. First: for something as important as production, access to untrusted actors is never granted (we don’t rely on asking a junior engineer or AI nicely). Second, he never separated progress from development. In a later public conversation on LinkedIn, Lemkin, who holds an executive MBA from Stanford and a JD from Berkeley, admitted that He was not familiar with best practices Distribution development and production database.
The takeaway for business leaders is that standard software engineering best practices still apply. We should add at least the same safety barriers for AI as we do for junior engineers. Arguably, we should go beyond this and treat AI a little differently: there are reports that like Hall in Stanley Kubrick 2001: A Space OdysseyAI can try to do Break its sandbox environment To accomplish a task. With more vibe coding, having experienced engineers who understand how complex software systems work and can implement appropriate safeguards in the development process.
Tea hack
Sean Cook is the founder and CEO of Tea, a mobile application launched in 2023 designed to help women date safely. In the summer of 2025, they were “hacked”.": 72,000 photos, including 13,000 authentication photos and official ID photos, Leaked on public discussion forum 4 China. Worse, Tea’s own privacy policy promises that these images will "Deleted immediately" After users are authenticated, that is, they likely violated their own privacy policy.
I use “hack” in the air coat because the phenomenon is less likely to be the cleverness of attackers than the unavailability of defenders. In addition to violating its data policies, the app left the Firebase storage bucket unprotected, Exposing sensitive user data on the public Internet. It’s the digital equivalent of locking your front door but leaving your back open with your family jewels hanging on the doorknob.
While we don’t know whether the root cause was coding or not, the Tea hack has highlighted catastrophic breaches stemming from basic, preventable security errors due to poor development processes. This is the type of vulnerability that is addressed through a disciplined and thoughtful engineering process. Unfortunately, the relentless push of financial pressures, where a “lean”, “fast and break things” culture is the polar opposite, and vibe coding only exacerbates the problem.
How to Safely Adopt AI Coding Agents?
So how should enterprise and technology leaders think about AI? First, this is not a call to abandon AI for coding. An MIT Sloan study Estimated AI gains between 8% and 39% in productivity, while a A McKinsey study 10% to 50% reduction in time to complete tasks with the use of AI.
However, we should be aware of the risks. The old lessons of software engineering don’t go away. These include many tried-and-true best practices, such as version control, automated unit and integration tests, security checks such as CAST/DAST, separating development and production environments, code review and secrets management. If anything, they become more prominent.
AI can produce code 100 times faster than humans, providing an illusion of productivity that is an attractive siren call to many executives. However, the quality of the rapidly emerging AI sloop is still up for debate. To develop complex production systems, businesses need the thoughtful, hands-on experience of human engineers.
Tianui Michael Lee is president at the Pragmatic Institute and founder and president of the Data Incubator.
Read more from us Guest authors. Or, consider submitting a post of your own! See our Guidelines here.