Table of contents
Happy January, all.
I hope you had a restful winter break and are ready for an exciting 2024.
As we kick off a new year, I wanted to write this note to highlight some of the trends I’m particularly excited about regarding GenAI code in the software development lifecycle. I also want to share some of the work we’re doing to help our customers stay compliant with emerging global regulations.
But before I jump in, I’d like to express how grateful I am, on behalf of our team, for the support and partnership of our advisors, customers, and investors. The collective worth of the codebases we assessed has now exceeded $1T across industries ranging from manufacturing to healthtech. We’ve come so far in building one of the most comprehensive codebase datasets on the market and we’re just beginning our journey. You helped us get here.
I hope you find these insights I’m sharing with you today helpful. As always, feel free to respond with questions, thoughts, and feedback.
Trends for Technical Due Diligence
There are three trends we’re following closely: overall market conditions, the continued growth of quantitative scans to supplement tech diligences, and a prediction for the rapid rise of GenAI code assessment as part of diligence.
- First, overall market conditions. As has been widely reported, market participants saw a substantial decline in M&A activity vs. 2022 heights.
Sema sees a light at the end of the tunnel. Not only do we agree with the analysts who predict a thawing by mid-year, but our own transaction and pipeline volumes indicate that the thawing has already begun.
- Second, growth in the use of quantitative scans. The 10,000 foot view shows an unmistakable trend towards quantitative scans becoming more and more essential for investor diligence.
Ten years ago, quantitative scans were rare. Five years ago, we saw the rise of quantitative scans to manage Open Source risk in particular. Within the past year, quantitative scans for Open Source risk have become nearly mandatory, and comprehensive scans covering Open Source plus security, team, etc. have become the rule rather than the exception.
- Third, there’s significant investor interest in measuring GenAI code usage and risk. Based on conversations with industry-leading investors, operators, and advisors this fall and early winter, the demand for GenAI code as a module of diligence code scans is nearly unanimous.
Announcing Sema’s AI Code Monitor
We expect by the end of 2024 that GenAI code will be as widespread as Open Source. The faster companies start responding to this paradigm shift systematically and at scale, the faster they will reap the rewards and avoid the significant risks.
At the request of several of our clients, Sema is building a new product called the AI Code Monitor, which helps technical leaders, investors, acquirers, and compliance officials:
- Understand the risks of investing in/acquiring tech and tech-enabled businesses given their use of GenAI code in the Software Development Lifecycle (SDLC)
- Quantify the current level of developer productivity improvements using GenAI, and how to drive further gains
- Stay in compliance with the rapidly changing landscape of GenAI regulation affecting the SDLC
The research we did in launching this product highlighted the urgency of monitoring AI-generated code. As everyone knows, even though GenAI usage varies widely across and within organizations, usage is skyrocketing. And yet, growth notwithstanding, we are in the early stages of adoption of GenAI for code. As adoption increases, so does scrutiny of the technology and potential risks.
Compliance Standards for Generative AI for Code
One of the clearest takeaways from our research is how uncertain the regulators are on how exactly they will regulate GenAI code. The US Copyright Office, for example, acknowledged that "Questions remain about where and how to draw the line between human creation and AI-generated content."
Sema is investing heavily in this area to support our clients and partners.
- In Q3 2023, we built a Regulatory Affairs team to track and advise on GenAI for code regulation and legislation
- We are building what we believe is the most comprehensive, and most actionable, database of current and potential legislation, regulations, and guidance
- We launched a free newsletter series to stay up to date on the global regulatory landscape
“Actionable” is the key phrase for us. We are not only tracking the potential guidelines, but we are turning those results into simple rules that help companies see if they are on track or not, overall and by product, with real-time alerting. Thus far we are able to monitor over 250 potential rules based on 42 distinct sources of compliance guidance, at the state, national, and continental level.
How to Get Involved
Thank you again for being part of the Sema community – whether you’ve played a direct role in our journey or we only met or spoke briefly.
If you are interested in a demo of our AI Code Monitor or a trial of our compliance offerings, go ahead and reply to this note and I’ll set you up with a member of our team. I’d be just as delighted to hear from you with any feedback on these insights or to have a technical discussion about how you’re thinking about this new genAI-powered world.
Wishing you a wonderful and productive 2024.
All my best,
Matt
