Salesforce leans on AI for nearly half its operations: CEO

Salesforce CEO Marc Benioff said AI now handles 30% to 50% of tasks within the company, including software engineering and customer service.

He added that the adoption of AI has reduced the need for additional hiring.

This trend is similar to what’s happening at other tech firms like Microsoft and Alphabet.

Salesforce is developing an AI tool that can automate customer service tasks. The tool has reached 93% accuracy and is already in use by major clients, including Walt Disney Co.

As AI becomes more integrated into operations, Benioff noted a shift in employee roles, saying, “We can move on to do higher-value work.”

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ AI adoption follows historical labor transformation patterns rather than causing mass unemployment

Salesforce’s 30-50% AI integration reflects a broader historical pattern where technological advances reshape work rather than eliminate it entirely.

McKinsey research shows that while approximately 50% of current work activities are potentially automatable, less than 5% of jobs can be fully automated with existing technologies 1.

The International Labour Organization supports this view, finding only 2.3% of jobs globally are at high risk of complete automation, with essential roles being particularly resistant to replacement 2.

This pattern aligns with previous technological transitions. For example, the shift from agriculture to manufacturing created new job categories. Similarly, AI is creating roles in data analytics and AI ethics while transforming existing positions 3.

Salesforce demonstrates this transformation by reducing engineering and customer service hiring while simultaneously planning a 22% increase in sales roles, suggesting a reallocation rather than elimination of human labor 4.

2️⃣ Organizational AI adoption varies dramatically across functions with specific economic impacts

Benioff’s 30-50% figure represents an average across functions, with implementation rates varying significantly by department and task type.

McKinsey’s analysis reveals that the most automatable tasks are physical activities in predictable environments and data processing, while roles requiring human interaction face less disruption 5.