She saw the AI software collapse coming almost a year ago. Here's what she expects next.

• Michelle Miller and her colleagues at AlixPartners warned of trouble brewing back in April.
• Now, software stocks have plunged on AI disruption concerns that Miller laid out many months ago.
• Here's where Miller sees the software industry headed now.

Almost a year ago, Michelle Miller co-authored a prescient study that warned the software industry would be squeezed by the rise of generative AI.

Now, software stocks have plunged for many of the same reasons Miller and her colleagues at AlixPartners laid out back in April.

So, I checked back in with Miller this week to see where she sees things going from here. As co-lead of the Enterprise Software Practice at AlixPartners, she's steeped in this stuff, so pay attention.

Business Insider: Why have software and SaaS stocks slumped so much in recent weeks?

Michelle Miller: We think several factors are behind the slump in software / SaaS stocks in recent weeks, but we would also say that these forces have been pressuring the software industry for a while and are now accelerating. These include pent-up fear around AI disruption, outdated pricing models, general macro uncertainty, as well as concerns around AI-based valuation frameworks and valuation premiums. Previously, enterprise software company valuations were driven by investors' growth expectations. Now, software growth is being challenged by the AI-driven forces I mentioned earlier, and we see multiples decreasing.

Regarding trending news (e.g., Anthropic's Cowork), this is another front-of-mind example of an AI tool lowering the barrier to entry, gaining traction, and disrupting incumbent workflows.

Can you boil this threat down to a simple idea that non-technical people can understand?

AI won't eliminate the need for software companies, but software companies will need to prove that they can deliver on their growth agenda.

Software companies will need to learn how to navigate the new era of fundamentally different business economics. AI is forcing change across software development, AI governance and data security, go-to-market operations, pricing models, valuation frameworks, and business structure. We believe that software companies mastering these transitions will define the winners in the next era, and those unable to adapt will be sidelined as the industry fundamentals are redrawn.

Is there a broader meaning beyond the strange Moltbot chats online? What are the serious implications of this tool (now called OpenClaw)?

AI continues to capture the imagination in new ways and push the bounds of progress. Regarding Moltbot and Moltbook, these are the latest entrants to validate this imagination capture, showcasing the remarkable speed of progress of AI technology.

I see these as experiments that have scaled and have received significant attention, particularly in AI communities. They demonstrate the very early innings of agent-to-agent interaction, highlighting the productive power single users will likely have at their fingertips with modern agentic systems. This raises serious questions around AI governance and data security for enterprises dealing with concerns around open-source and shadow AI. The so-called "trust infrastructure" has remained an afterthought rather than a foundational capability, making it a critical barrier to broader AI adoption.

How are companies changing how they use software? Are they using AI tools instead in some situations? Can you explain an impactful example?

Companies are piloting AI tools across the organization. In product and engineering, companies are leveraging AI-accelerated coding, creating conversational interfaces, and developing trust infrastructures. In go-to-market functions, companies are implementing AI sales tools and new pricing models. Companies are also using AI to enhance general administrative functions, such as finance, HR, and corporate IT process automation.

More than 30% of tech company workflows already include AI tools today, and that is expected to increase substantially over the next 5 years. However, a majority of GenAI proofs of concept are not moving into production, with 75% of all AI deployments expected to fail. Winning companies in 2026 will stop running incremental pilots and start reinventing how work gets done.

Which areas of the software and SaaS industry are most exposed to this disruption and why?

All software/SaaS industry sectors are exposed to AI disruption. We expect that no aspect of the enterprise software business escapes unscathed. Most dramatically, many mid-market software companies face an imminent existential threat, leading to significant consolidation. We predict that in 2026, AI disruption will force major consolidation in the mid-market enterprise software industry, with M&A deal volume increasing 30-40% YoY as the middle market is being squeezed from slower growth, AI startup investment, and tech giants' bets/investments.

Are there any existing software/SaaS services that will remain essential and useful, and therefore immune to AI disruption?

Regarding mid-market, general-purpose productivity and workflow automation software will be hit hardest. Incumbents with proprietary data moats, platform entrenchment, and vertical specialization in regulated industries, like HIPAA-compliant patient data platforms, will fare better. Vendors that evolve their offerings to help customers transform static data sources into actionable context will fare best.

Sign up for Business Insider's Tech Memo newsletter here. Reach out to me via email at abarr@businessinsider.com.