Ready or not, enterprises are betting on AI (techcrunch.com)

This week saw a flurry of enterprise AI deals that signal companies are shifting money and attention from speculative consumer apps to pragmatic business use cases. Zendesk unveiled AI agents it says can resolve roughly 80% of customer-service issues, Anthropic announced strategic partnerships with IBM and Deloitte, and Google launched a new AI-for-business platform — all moves that push large language models (LLMs) and voice/text automation into core enterprise workflows like service desks, reports, and internal tooling. The rush to deploy LLMs at scale underscores two big implications for AI/ML teams: opportunity and risk. On the upside, enterprise deployments promise near-term revenue and measurable productivity gains by automating email/voice/text triage and creating end-to-end service suites. On the downside, a recent incident where a Deloitte-delivered report appeared to contain AI-generated hallucinations — and prompted government pushback — highlights persistent model failure modes, the need for human-in-the-loop verification, strong provenance and audit trails, and governance around training data and outputs. Technical teams will need robust evaluation, monitoring, and guardrails (source validation, citation checking, uncertainty estimation) to translate these high-profile deals into reliable, compliant systems.