Salesforce CEO Marc Benioff says the company has cut 4,000 customer support staff for AI agents so far
The Salesforce CEO boasted of “rebalancing headcount” in support roles
The jury may still be out on whether generative AI is going to cause widespread job losses, but the impact of the technology is already being felt at Salesforce.
Speaking on The Logan Bartlett Show, CEO Marc Benioff claimed the CRM company has cut its customer support workforce by almost 50% thanks to the use of its own agentic AI platform, Agentforce.
“We’re customer zero for our new agentic service and support product at Salesforce,” Benioff said. “We have now done about a million and a half conversations with our customers (through the agentic layer) and at the same time … a million and half conversations also happened through our support agents during that same period.
“The SEASAT scores were about the same, which was stunning and also I was able to rebalance my headcount on my support – I’ve reduced it from 9,000 heads to about 5,000 because I need less heads.”
Despite these cuts, Benioff said the way the remaining humans are working with the AI agents - under the guidance of an omnichannel supervisor - “is the most exciting thing that’s happened in the last nine months for Salesforce”.
Agentforce was launched at Salesforce's annual company conference, Dreamforce, in 2024. Since then, Benioff has pitched the service as an answer to demographic changes such as global declining birthrates and labor shortages, which will make it harder to find and recruit new talent.
In a previous interview with Logan Bartlett, he mused that the company may not hire any new engineers in 2025 as it had seen “such incredible productivity gains because of the agents that work side-by-side with our engineers”.
Sign up today and you will receive a free copy of our Future Focus 2025 report - the leading guidance on AI, cybersecurity and other IT challenges as per 700+ senior executives
Benioff noted that cost cutting wouldn’t be the main benefit of AI to businesses, but that it could also lead to an increase in revenue.
The Salesforce chief claimed that in the 26 years the company has been in operation it had over 100 million potential new client leads it hadn’t called back because there weren’t enough staff to do so.
Agentic AI, however, will supplement the human workforce and enable the company to follow up on all leads, which he claimed number more than 10,000 a week.
Expect more on agents at Dreamforce 2025
Benioff’s conversation with Barlett comes a little over a month before the company’s annual Dreamforce event in San Francisco.
While he gave away little in terms of what will be featured – other than the entertainment for the final evening – given the tone of the conversation, including claims that people entering the workforce now are “AI natives [and] agentic natives”, it’s likely to focus heavily on AI agents once again.
In recent weeks the company has launched Agentforce for Public Sector, highlighted its “advancing AI agent training” with enterprise digital twins at Salesforce AI Research, and released a blog titled “In the agentic era, human imagination must shape the jobs of tomorrow”.
In May this year, it also announced a planned acquisition of Informatica to enhance its agentic AI service – something that may well also be discussed during the event.
ITPro will be reporting live from Dreamforce, which runs from 14-16 October, so be sure to check back then for news, updates, and exclusives from the event.
Make sure to follow ITPro on Google News to keep tabs on all our latest news, analysis, and reviews.
MORE FROM ITPRO
- IT leaders don’t trust AI agents yet – and they’re missing out on huge financial gains
- Marc Benioff thinks today’s CEOs will be the last to have a fully human workforce
- Marc Benioff warns Microsoft could repeat 'pretty nasty’ Slack playbook with OpenAI

Jane McCallion is Managing Editor of ITPro and ChannelPro, specializing in data centers, enterprise IT infrastructure, and cybersecurity. Before becoming Managing Editor, she held the role of Deputy Editor and, prior to that, Features Editor, managing a pool of freelance and internal writers, while continuing to specialize in enterprise IT infrastructure, and business strategy.
Prior to joining ITPro, Jane was a freelance business journalist writing as both Jane McCallion and Jane Bordenave for titles such as European CEO, World Finance, and Business Excellence Magazine.
-
Marc Benioff says hiring in software engineering is ‘mostly flat’ at Salesforce because of AINews Salesforce CEO Marc Benioff has revealed hiring for software engineering has dipped as a result of AI, but the CRM giant is ramping up recruitment in other key areas to push its agentic agenda.
-
Are AI browsers a golden opportunity or cybersecurity nightmare?In-depth AI browsers are on the rise despite the concrete risks associated with using them
-
The AI-enabled Slackbot is now generally available – Salesforce says it could save more than a day’s work per weekNews With an entirely overhauled model behind the chatbot, users can summarize channels and ask for highly personalized information
-
Retailers are turning to AI to streamline supply chains and customer experience – and open source options are proving highly popularNews Companies are moving AI projects from pilot to production across the board, with a focus on open-source models and software, as well as agentic and physical AI
-
Microsoft CEO Satya Nadella wants an end to the term ‘AI slop’ and says 2026 will be a ‘pivotal year’ for the technology – but enterprises still need to iron out key lingering issuesNews Microsoft CEO Satya Nadella might want the term "AI slop" shelved in 2026, but businesses will still be dealing with increasing output problems and poor returns.
-
OpenAI says prompt injection attacks are a serious threat for AI browsers – and it’s a problem that’s ‘unlikely to ever be fully solved'News OpenAI details efforts to protect ChatGPT Atlas against prompt injection attacks
-
OpenAI says GPT-5.2-Codex is its ‘most advanced agentic coding model yet’ – here’s what developers and cyber teams can expectNews GPT-5.2 Codex is available immediately for paid ChatGPT users and API access will be rolled out in “coming weeks”
-
Google DeepMind CEO Demis Hassabis thinks startups are in the midst of an 'AI bubble'News AI startups raising huge rounds fresh out the traps are a cause for concern, according to Hassabis
-
OpenAI turns to red teamers to prevent malicious ChatGPT use as company warns future models could pose 'high' security riskNews The ChatGPT maker wants to keep defenders ahead of attackers when it comes to AI security tools
-
AWS has dived headfirst into the agentic AI hype cycle, but old tricks will help it chart new watersOpinion While AWS has jumped on the agentic AI hype train, its reputation as a no-nonsense, reliable cloud provider will pay dividends