SaaS valuations: how AI is reshaping software investment

Generative AI is lowering the code barrier that code barriers that historically provided software-as-a-service (SaaS) companies with a strong competitive moat. At Luxembourg Venture Days, leading investors outlined code alone no longer shields startups from competition. On 15 October 2026, a panel, moderated by Amélie Madinier, Executive Director at Aureus Ars & Scientia, explored how the erosion of traditional software moats is shifting investor focus to proprietary data and workflow ownership.

Why code is no longer a lasting competitive moat for SaaS valuations

Historically, software-as-a-service was considered a cornerstone of venture capital portfolios because of its scalable business models, predictable recurring revenues and high margins. However, the rise of generative AI has disrupted these traditional valuation models. Thibaut Schlaeppi, Managing Partner of RAISE Ventures, explained that the barrier to entry has shifted. "What we usually call the barrier to entry used to be the code, but it no longer is. Because codes get generated overnight now," Mr Schlaeppi said.

This collapse of the code barrier means that competition has intensified. Anyone can launch a software application with fewer resources, leading to higher customer acquisition costs and shorter customer lifetimes. Sébastien Paillet, Founder and CEO of ScaleX Invest, noted that SaaS multiples are currently undervalued but that opportunities remain for startups that manage their cash flow effectively. "The core of the business model and the capacity to generate cash still hold value for SaaS," Mr Paillet said. The question, he added, is now more about growth, traction and the revenue model, all of which AI is disrupting.

What investors look for now: data, workflow and team

To maintain strong SaaS valuations, investors now look beyond technical code. Cristian Negrutiu, Founding Partner of Sparking Capital, emphasised that venture capitalists do not invest in technical ideas alone. "We as investors don't put money into technical ideas. We invest in businesses," Mr Negrutiu said. He advised startups to find a specific niche and build something genuinely useful for their customers. If a startup owns the workflow and has proprietary data, it maintains a distinct advantage.

This shift is also transforming how venture capital firms evaluate founding teams. Investors are increasingly backing teams led by mathematicians, researchers and PhDs who understand the core mechanics of AI. However, technical expertise must be balanced with commercial capability. Mr Negrutiu warned that some highly technical founders lack the ability to scale their businesses. He stressed the importance of having a salesperson or a product person in the team. To address this, RAISE Ventures has introduced a new evaluation method using a third-party partner to assess leadership and team risks objectively.

As the market moves past the initial wave of foundational models, the focus of AI disruption is shifting toward real-world applications in regulated industries. A prime example is RAISE Ventures’ recent investment in Fleuret AI, a French startup automating penetration testing using AI agents.

How Luxembourg AI Factory supports software companies

To navigate these valuation shifts, Luxembourg-based software companies can leverage targeted support from Luxembourg AI Factory. For instance, the Data Exploration & Storytelling service helps startups structure proprietary datasets to to build sustainable competitive advantages. Additionally, the AI Solution Scouting and Assessment service helps resolve the "buy vs. build" dilemma by validating proofs of concept. Technical Deep Assessment also provides objective architectural reviews to ensure long-term scalability. Software businesses can explore these pathways on the Luxembourg AI Factory website.