Enterprise shift to agents
Enterprise AI is moving from one-off chatbots to 'agentic workflows' — OpenAI says enterprise customers now make up over 40% of its revenue as firms assemble teams of specialised agents. Vendors are responding with governance and integration features, and companies are building full data-and-AI platforms with cloud and compliance partners to embed personalised, governed agent workflows. ((decrypt.co), Nutanix Bets on Agentic AI Governance, Minor Hotels Unveils Global Data and AI Platform)
A year ago, most companies were still bolting one chat window onto a website or an employee portal. This week, OpenAI said business customers now generate more than 40% of its revenue, and the company expects enterprise revenue to reach parity with consumer revenue by the end of 2026. (openai.com) The shift is from one helper to a small digital staff. Instead of asking one chatbot to do everything, companies are stitching together specialized agents for coding, search, customer support, approvals, and data lookups inside the same workflow. (decrypt.co) OpenAI put numbers on how fast that change is moving. It said its application programming interfaces now process more than 15 billion tokens per minute, Codex has reached 3 million weekly active users, and customers now include Goldman Sachs, State Farm, DoorDash, Thermo Fisher, Philips, Cursor, and LY Corporation. (openai.com) That sounds like a software upgrade, but it behaves more like an operations redesign. A single chatbot answers a question, while an agent workflow can pull a customer record, check a policy, draft a reply, send it for approval, and log the action in another system. (decrypt.co) Once companies let software take several steps instead of one, the hard part stops being the model and becomes the guardrails. Nutanix used its.NEXT conference this week to pitch new tools for governing agentic artificial intelligence across hybrid and multicloud systems, with controls aimed at optimization, security, and oversight. (markets.businessinsider.com) Nutanix is betting that big companies will not run all of these agents in one place. Its pitch is that banks, manufacturers, and healthcare groups will need agent systems that move across private data centers and public clouds without losing policy controls or visibility. (siliconangle.com) The hotel industry is now building around the same idea from the customer side. Minor Hotels said on April 10 that it is creating a global data and artificial intelligence platform with Google Cloud, Salesforce, OneTrust, and Deloitte to connect guest data across its brands and use that data for personalized service and marketing. (ryt9.com) Minor operates more than 560 hotels, resorts, and residences in 58 countries, so this is not a pilot tucked inside one app. The company described the project as one of the most significant technology investments in its history and said it is building the platform from the ground up rather than layering it onto old systems. (skift.com) OneTrust’s role in that stack is a clue to where this market is going. Personalization only works if companies can connect identity, consent, and customer data cleanly enough that an agent can act without crossing legal or privacy lines. (ryt9.com) So the new enterprise race is not just who has the smartest model. It is who can tie models to company data, plug them into existing software, track every action, and prove to a compliance team that the system did exactly what it was allowed to do. (decrypt.co, siliconangle.com, ryt9.com)
