Making Speed Safe¶
You Have Heard These Concerns Before¶
When teams in government software organizations first started adopting agile delivery methodologies, people who were unfamiliar with this way of working had a predictable reaction: slow down.
The objections sounded like this:
- "How can it be secure if teams are moving that fast?"
- "We need more oversight, not less."
- "This works for startups, not for us."
When development teams first started shipping on a weekly basis or faster, there were real concerns about security, quality, and whether speed was being prioritized over safety. Those concerns were valid. But the answer was not to slow down. The answer was to put practices in place around the software development lifecycle that made it safe to go fast.
That is the mindset shift. The best technology organizations in the world do not move slowly. They move quickly, with discipline. And that discipline comes from practices.
The Practices That Matter¶
If you have been around agile transformation, DevSecOps adoption, or any modern approach to software delivery, these will sound familiar:
| Practice | What It Does |
|---|---|
| Continuous integration and continuous delivery (CI/CD) | Code changes are automatically built, tested, and prepared for release on a continuous basis |
| Automated security scanning | Every change is scanned for vulnerabilities before it ships |
| Comprehensive testing | Automated tests verify that software works as expected before it reaches users |
| Infrastructure automation | Environments are provisioned and managed through code, not manual processes |
When these practices are in place, speed is not reckless. It is controlled. That is how the best organizations operate, and it is what gives them their competitive edge.
AI Dials All of This Up¶
AI-generated code still needs to be tested. It still needs to pass security scans. It still needs to be reviewed. The fundamentals do not change; they just matter more when the volume and velocity go up.
The pattern is the same one you saw with agile delivery: a powerful new capability arrives, it disrupts established workflows, and the first instinct is to slow it down rather than learn how to steer it. But the organizations that win are the ones that put practices in place, not brakes.
There are also new practices that come with responsible AI development: how you instruct AI, how you give it context about your domain and your standards, how you verify its output. Those practices are what this lab is about.
What This Lab Will Give You¶
By the end of this lab, you will have:
- Firsthand experience building with AI. You will describe what you want in plain English and watch AI build it.
- An understanding of the practices that make it safe. You will learn how AI works, why it produces variable results, and what practices keep that variability under control.
- The vocabulary to guide your organization. You will leave with the understanding to have informed conversations about AI capabilities, risks, and the practices your teams need.
Round Robin: What Concerns Are You Hearing?
Team Discussion | ~3 minutes total | Round robin, one person at a time.
In a few words, share with your team: what concerns do you or others in your organization have about AI adoption?
Key Insight
The answer to speed is not to slow down. It is to put practices in place that make speed safe. The same lesson from agile delivery applies to AI, and the stakes are higher.
Before you can steer it, you need to understand how it works. Let's start there.