More than 50% of the software enterprises currently purchased could be replaced by artificial intelligence systems, according to Arthur Mensch, chief executive of Mistral AI, adding fuel to investor anxiety over the durability of traditional software business models.
Speaking to CNBC at the India Accelerates event on the sidelines of the AI Impact Summit in New Delhi, Mensch said: “I would say more than half of what’s currently being bought by IT in terms of SaaS is going to shift to AI.”
His comments come amid a sharp correction in software stocks, partly triggered by advances in enterprise-focused AI tools such as Anthropic’s Cowork product. Investors increasingly worry that generative AI systems can replicate — or outperform — many of the functions currently delivered by subscription-based software platforms.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
The iShares Expanded Tech-Software Sector ETF, which includes major holdings such as Microsoft and Salesforce, has fallen more than 20% this year. In India, software giants such as Tata Consultancy Services and Infosys have also declined, reflecting broader concerns that AI may compress demand for legacy enterprise solutions.
The “Replatforming” Argument
Mensch framed the shift not as incremental substitution but as structural replatforming. He argued that AI systems can now generate custom applications within days — replacing what once required specialized vertical SaaS vendors.
“AI is making us able to develop software at the speed of light,” he said.
According to Mensch, when enterprises have the appropriate infrastructure — particularly the ability to connect internal data securely to AI systems — they can rapidly build workflow applications for procurement, supply chain management, and other operational processes.
“Five years ago, you would actually need a vertical SaaS,” he said. Now, AI tools can create tailored workflow software in a matter of days.
The implication is that companies may bypass traditional software vendors entirely, instead using foundation models and AI development platforms to build internal tools.
Mensch described a “replatforming” trend in which enterprises reassess decades-old IT systems. He said Mistral now has more than 100 enterprise customers exploring the replacement of older software stacks, particularly where licensing costs have escalated.
“They see AI as a way to replatform the thing so that it becomes more efficient and less costly,” he said.
Mensch drew a distinction between workflow software — which he sees as vulnerable — and systems of record, which he believes will remain foundational.
Systems of record manage core enterprise data, including financial ledgers, HR records, and customer databases. These platforms often serve as the backbone of corporate IT architecture and are deeply embedded in compliance and operational processes.
“Systems of records are not going to change,” Mensch said, suggesting AI will sit on top of these databases rather than replace them.
Bipul Sinha, CEO of Rubrik, echoed that view in a separate CNBC interview, arguing that workflow software could face significant disruption, while data infrastructure enabling AI could benefit from increased demand.
This distinction is central to investor strategy. If AI primarily replaces user-interface-driven workflow tools while reinforcing demand for data storage, governance, and cloud infrastructure, then the winners and losers within the software sector may diverge sharply.
Market Implications for SaaS Models
The software-as-a-service model has long relied on predictable subscription revenue, high gross margins, and long-term contracts. If enterprises shift from buying packaged software to building AI-driven custom applications, that revenue model could come under pressure.
Instead of paying recurring fees for standardized tools, companies might invest in AI infrastructure and in-house development talent. Over time, this could reduce vendor lock-in and increase price sensitivity.
The speed of AI-driven application development also challenges the traditional sales cycle. If a company can prototype and deploy an internal workflow tool in days, it may be less inclined to negotiate multi-year contracts with external SaaS providers.
Investors are grappling with whether AI represents incremental enhancement — embedded within existing platforms — or a fundamental threat to the SaaS ecosystem.
Mistral’s India Expansion
Mensch also outlined plans for Mistral AI to open its first office in India this year. The move signals the company’s intent to compete in one of the fastest-growing AI markets globally.
While Mistral is building its own data centers in Europe, its India strategy will rely on partnerships with firms that already operate local infrastructure. This approach reflects regulatory sensitivities around data sovereignty, as India encourages AI providers to ensure domestic data storage and local processing capabilities.
Mensch said Mistral is already working with multinational firms that have operations in India and is now actively pursuing public- and private-sector customers based in the country.
India’s linguistic diversity, including languages such as Hindi and Punjabi, presents both a technical challenge and a commercial opportunity. Mistral’s large language models are designed to accommodate multilingual inputs, which Mensch described as crucial for long-term consumer adoption.
“That’s something that down the line will be super important for the Indian consumer market,” he said.
Mistral, founded in France, positions itself as a European alternative to U.S.-based AI giants such as Anthropic and OpenAI. Its expansion into India underscores the intensifying global competition for enterprise AI customers.
As enterprises weigh whether to integrate AI into existing SaaS platforms or rearchitect systems around AI-native tools, the debate over software displacement is likely to intensify.
Mensch’s assertion that more than half of enterprise SaaS spending could shift to AI may be contested. But the underlying trend, rapid AI-driven customization, workflow automation, and infrastructure investment, is reshaping how investors evaluate software companies.