Top executives at Salesforce have publicly admitted that they overestimated AI’s readiness for real-world deploymes and rushed to replace humans.
According to Maarthandam, Salesforce’s aggressive push to automate customer-related operations and the widespread use of artificial ielligence is now eering a phase of public accouability. The company has now officially admitted that firing more than 4,000 experienced employees and replacing them with artificial ielligence was a wrong decision.
Salesforce is one of the pioneers in the use of artificial ielligence
This admission came in the wake of Salesforce’s decision to eliminate about 4,000 customer support jobs in 2025. Following this action, Salesforce reduced its support force from about 9,000 people to nearly 5,000 people. Initially, manageme had touted the layoff as a success in increasing productivity thanks to age-driven artificial ielligence systems capable of managing customer conversations at scale.
But top Salesforce executives have acknowledged in rece iernal conversations and public commes that the company has been overly optimistic about the ability of AI systems to fully replace human judgme, especially in complex customer service scenarios.
According to the company’s executives, the automated systems struggled to deal with subtle issues, cases requiring referrals to higher levels, and infreque but complex customer problems. The result was a decline in service quality, an increase in the volume of complais, and emergency iernal efforts to stabilize operations that were previously managed by experienced staff.
Salesforce CEO Marc Benioff, who previously touted AI as a transformative force and made it possible for the company to operate with fewer human resources, has since softened his tone and somewhat backtracked on his previous positions.
