Overview
- Investors sold first and asked questions later after new agentic tools surfaced use cases, with Altruist’s Hazel sending Schwab, Raymond James, and LPL down 7% or more on Feb. 10 and earlier Insurify news hitting insurance brokers, as Goldman Sachs pegged last week’s value loss near $2 trillion and volatility persisted.
- SaaS platforms are unlikely to vanish as many agents use them rather than replace them, yet Anthropic’s Claude Cowork and OpenAI’s enterprise agent platforms threaten to capture the orchestration layer, even as Nvidia’s Jensen Huang called the recent ‘SaaSpocalypse’ narrative illogical.
- Enterprise vendors are retooling revenue models to defend growth: Salesforce is testing an Agentic Enterprise License Agreement for all‑you‑can‑use Agentforce, ServiceNow is shifting toward consumption and value‑based pricing, and Microsoft has added consumption elements to Copilot Studio.
- Early labor effects are showing up in data and corporate actions, including a Stanford finding of a roughly 13% drop in early‑career employment and about 55,000 AI‑related layoffs in 2025, with Microsoft cutting 9,000 jobs and Salesforce eliminating 4,000 customer service roles.
- Scaling constraints and governance worries temper the hype, as experts flag hallucination‑driven error cascades when chaining agents, Stuart Russell estimates a 75% chance current AGI efforts fail and warns success would be dangerous, and Tristan Harris criticizes a safety‑cutting race as federal rules were rolled back even as 26 states advanced AI laws.