AI Agents Reshape Workflows, Create New Opportunities

Organisations will soon see every employee become a "boss" managing personalised teams of artificial intelligence agents, according to OpenText, an information management company. This vision suggests AI agents will automate processes, significantly boosting human accomplishment and efficiency. OpenText reports that 99 of the 100 largest global companies are its customers.

"There will be hundreds and perhaps thousands of bots being created by everybody in this audience," OpenText Chairman and Chief Strategy Officer Tom Jenkins said at a company conference. He added that "everyone is going to be building their own little AIs.”

Numerous companies are already developing AI agents. Amazon uses them to find bugs, while customer engagement platform Braze is building agents. Startups create AI agents for investment bankers, and Hubspot, Adobe, and Salesforce are also developing them.

Salesforce reported that AI agent use exploded 22 times this year. OpenText is launching a new AI data platform to power secure, contextual agents, alongside an agent-building platform and an agent-to-agent integration platform for business workflows.

The proliferation of these agents will necessitate new management systems, according to Savinay Berry, OpenText Chief Technology Officer and Chief Product Officer. He stated that "agent lifecycle management" will become essential, covering creation, governance, management, and retirement, similar to human lifecycle management.

AI agents are gaining popularity due to their ability to provide time savings, cost reductions, and "superpowers" for human workers. The ultimate goal is to achieve more at a lower cost.

OpenText experienced an 80% time savings by automating release notes with agents. Customer United Airlines saw an 85% speed-up in priority one software problem resolution, reducing it from 30 minutes to four minutes in one instance.

Another OpenText customer, IBM, reported that AI agents now handle 94% of HR issues for its nearly 300,000 global employees. Honda saves thousands of hours monthly on inventory and production line issues with AI agents.

Pick n Pay, a South African retailer, now ships software with 20% to 30% fewer bugs thanks to AI agents. These substantial efficiency gains highlight the clear return on investment (ROI) driving companies to adopt AI agents rapidly.

Salesforce Executive Vice President Kishan Chetan noted that a single flight cancellation can generate hundreds of urgent rebooking needs. While a human agent might handle a dozen complex calls per hour, agentic AI can scale to automatically manage tens of thousands of customer conversations simultaneously.

Chetan emphasised that agentic AI is becoming an "absolute requirement" for delivering customer satisfaction, especially during mass disruptions. These disruptions, caused by worker shortages or major storms, strain systems and create ripple effects for weeks.

The increased volume and complex inquiries during such times place immense pressure on human agents. Chetan explained that the focus is shifting from scaling human teams to scaling human capacity with digital labour.

However, the widespread deployment of AI agents presents challenges, particularly regarding security. Berry warned that "the attack surface is going to completely increase orders of magnitude." He elaborated that the expanded attack surface is no longer just humans, but software itself.

AI-generated fraud is surging, with dark-web posts on using AI agents for payment fraud and malicious transactions reportedly increasing by 450%, according to Visa. This is concerning because AI agents performing significant tasks require substantial access to corporate systems, similar to human access.

Improperly configured agents could theoretically be prompted to extract sensitive company data. These security issues require resolution before companies deploy agents at scale.

Another concern is the potential impact on human employment. Berry, however, pushed back against the narrative that AI will eliminate most jobs or programmers. He believes organisation sizes will remain similar, but the work will be more effective.

Berry stated that skill profiles will change, with middle managers, who may be less fluent with technology and not yet senior enough for strategy, facing the highest risk. He stressed that computer science majors provide essential "systems thinking processes" crucial for managing complex agentic workflows.

Knowledge work will evolve to combine systems thinking with AI orchestration. Individuals may manage four or five key AI agents, which will, in turn, interact with additional agents to complete tasks.

Berry projected a "massive reskilling movement" similar to the transition from an agricultural to an industrial economy in the early 1900s. This transformation will require new skills, and daily work is likely to look significantly different.

---

• OpenText envisions every employee managing personalised AI agent teams, automating processes for increased efficiency.

• AI agents offer significant time and cost savings, with examples like an 80% time reduction for OpenText’s release notes and IBM’s 94% HR issue resolution.

• Security concerns, including an expanded "attack surface" and surging AI-generated fraud, pose challenges due to agents needing extensive access to corporate systems.

Source: FORBES