SaaS Business Leader Warns “The Old Moat Is Gone” After Rebuilding 20 Years of Software in 3 Days. Here’s What Still Protects Software Companies From AI
A SaaS executive rebuilt two decades of software in three days using AI, and his conclusion about what that means for the entire industry is unsettling for investors who think strong products are enough protection.
A recent Motley Fool Money segment addressed a growing fear among software investors: that generative AI could make much of the SaaS industry obsolete. Adam Field, the Chief AI Officer at Tungsten Automation, argued the fear is half right. One kind of software moat has effectively been destroyed while another kind of moat has become stronger than ever.
Rebuilding 20 Years of Software in 3 Days
Adam Field noted that the worry is that if a language model can generate working software on demand, why would enterprises keep paying subscription fees for tools that a small team, or an AI agent, could rebuild in a weekend?
His own experience made these fears feel real. He described rewriting a 20-year-old personal productivity app in three days using Cursor, with data migration so fast he “went to go make a tea and came back and it was done.” If a founder can rebuild two decades of code in a long weekend, the old assumption that switching costs and accumulated features keep customers locked in starts to look shaky.
That’s what led Field to say: “The old moat is gone.” The traditional SaaS advantage of being a “system of record,” where a company’s data lives and where workflows are cemented, is being eroded by tools that can replicate schemas, port data, and reconstruct functionality at a pace that was unimaginable even two years ago.
The SaaS Moats AI Still Can’t Replicate
The host pushed back hard on the doom case for one specific reason: not all moats are made of code. His example was invoice processing. Obtaining the licenses to process invoices in 140 countries, he said, “would cost millions of dollars and take years to do.” That is regulatory capital, not software capital, and a model cannot generate it.
He extended the point to expertise. Thirty to forty years of industry know-how cannot be replicated by a language model overnight. The durable SaaS moat, in his framing, is the combination of deep domain expertise and genuine risk transfer, where the vendor is legally, operationally, or financially on the hook if something goes wrong.
Software’s Next Winners Will Sell Expertise, Not Just Code
Generative AI is making software easier to build and customer data easier to move, weakening the traditional SaaS moat. The companies best positioned to survive are those whose value comes from deep industry expertise, regulatory approvals, and responsibility for customer outcomes, not just their code or functionality.
Contact [email protected] for any questions or corrections.
Thomas Richmond is a financial writer and content strategist with 5+ years of experience covering stocks and financial markets. He has published over 250 articles focused on individual stock analysis, helping investors better understand business fundamentals, stock valuations, and long-term opportunities.
Thomas previously served as a Content Lead at TIKR, a stock research platform, where he helped scale the company’s blog to hundreds of articles per month and contributed to a weekly newsletter reaching more than 100,000 investors.
He specializes in breaking down complex companies into clear, actionable insights for everyday investors, with a focus on fundamentals-driven research.
His work has also been featured on platforms including Seeking Alpha and Sure Dividend.
Outside of work, Thomas enjoys weight lifting and soccer.
