OpenAI has officially launched Presence, a new enterprise-grade service designed to deploy autonomous voice and chat agents capable of resolving customer and employee requests. The announcement signals a significant shift in how large organizations might handle frontline support, moving beyond simple chatbots toward agents that can operate IT systems, verify identities, and execute approved actions with minimal human intervention.
According to OpenAI’s official documentation, Presence allows enterprises to define strict boundaries for these agents. Companies can dictate exactly which actions an agent is permitted to take and when it must pause to seek human approval or transfer a case to a live representative. This governance layer is critical for industries with heavy compliance requirements, ensuring that automation does not bypass necessary regulatory checks.
OpenAI is not just selling the technology; it is using Presence internally to handle its English-language phone support channel. The system verifies callers, accesses account information, and completes approved actions autonomously. The company reports that Presence currently resolves 75% of inbound issues without human assistance. To further optimize performance, OpenAI integrates its Codex service, which monitors agent behavior and suggests process improvements. In internal tests, Codex’s recommendations reduced the need for human handoffs by 15 percentage points over a 10-day period.
Each Presence deployment is purpose-built for a specific task category, such as billing inquiries, insurance claims, or internal IT service requests. Agents receive only the knowledge base and system access required for that singular function, reducing the risk of errors or unauthorized data access. The platform also includes simulation and evaluation tools, allowing companies to stress-test an agent against company policy and operational goals before it ever touches a real customer.
Several major enterprises are already evaluating or piloting the service. Spanish bank BBVA is testing Presence for everyday banking support in Mexico, while Japanese technology group SoftBank is running trials focused on Japanese-language customer interactions. Australian insurer IAG is assessing whether the technology can help it manage surges in customer demand during severe weather events, a use case that highlights the system’s scalability during peak loads.
The rollout is not a self-service product. Enterprises must apply for a limited availability program, and integration is performed directly by OpenAI or selected global systems integrators. This managed approach ensures that complex enterprise environments receive the necessary configuration and support, though it may slow widespread adoption in the short term.
Workforce Impact and Job Security
While OpenAI’s announcement focuses on capability, it has sparked considerable debate regarding the future of support roles. The reported 75% automation rate raises immediate questions about staffing levels in customer service and IT help desks.
Pareekh Jain, CEO of Pareekh Consulting, advises CIOs to view the 75% figure as proof of concept rather than a guaranteed benchmark. He notes that OpenAI’s deployment benefits from being built around its own products and data. Large enterprises with fragmented legacy systems, uneven knowledge bases, and complex compliance demands will likely see lower initial automation rates. “Most organizations should expect lower initial automation levels that improve over time as the AI agent is refined,” Jain stated.
The consensus among analysts is that the initial workforce effect will be slower hiring rather than immediate mass layoffs. Tulika Sheel, senior vice president at Kadence International, explains that roles most exposed are repetitive, high-volume functions like frontline support and routine back-office processing. “However, I would expect the first impact to be on hiring and team growth rather than immediate large-scale job cuts,” Sheel said. She predicts that enterprises will eventually redesign roles around AI-assisted workflows, with humans focusing on complex cases, escalations, and relationship management.
Lian Jye Su, chief analyst at Omdia, adds that Presence is unlikely to drastically increase job displacement threats because enterprises have utilized similar customer-support automation from vendors like Genesys, NiCE, Five9, and AWS for years. She emphasizes that companies will use Presence alongside employees, with AI handling routine requests while people remain responsible for work requiring judgment and empathy.
Cost and Operational Risks
Beyond workforce concerns, CIOs must evaluate the financial and operational realities of deploying Presence. Analysts warn that maintaining resolution quality at scale is a significant challenge. “The key question is not simply how many tasks AI can handle, but whether it can handle them reliably at scale,” Sheel noted.
The financial case depends heavily on integration costs and ongoing governance. Jain points out that “often the biggest cost of enterprise AI is not tokens but integration and governance.” Companies will need to determine what systems and data agents can access, monitor performance continuously, and audit actions taken. These investments could offset early savings, particularly for organizations with outdated infrastructure.
Su agrees that enterprise IT complexity will prevent OpenAI from automating entire workflows in isolation. Enterprises will still need to collaborate with other technology providers and human employees. CIOs will favor systems that are fully auditable and seamlessly integrated with existing infrastructure. However, Jain suggests that economics could improve if companies reuse the same integrations and governance controls across multiple workflows, creating long-term efficiency gains.
What This Means for You
For IT leaders and business decision-makers, Presence represents a tangible step toward autonomous support operations. The managed rollout and focus on governance address many enterprise security concerns, but the path to full automation remains gradual. Organizations should start by identifying high-volume, repetitive tasks that are safe to delegate, while maintaining robust human oversight for complex or sensitive interactions. As the technology matures, expect to see broader availability and more competitive pricing, making enterprise AI agents a standard component of modern support strategies.
Source: Computerworld
Over to you: Are you ready to let AI handle your customer support, or do you prefer keeping humans in the loop for now?
You may also like
Apple’s MacBook Neo Strategy: How a Calculated Roadmap Is Reshaping the PC Market
Xbox PC Backwards Compatibility Runs on Xbox 360 Emulation, Opening the Door to 360 Games
Tech Layoffs 2026: A Month-by-Month Timeline of Workforce Cuts
NASCAR 26 Launches September 18 With Simultaneous PC Release and Major Gameplay Overhaul
