Why it Matters
The industry is witnessing the birth of a dangerous corporate trope: the ‘AI displacement fallacy’. Vendors are increasingly using AI as a sociopathic shield to mask fundamental governance failures, pandemic-era over-hiring, and mounting pressure on operating margins.
- Salesforce: Eliminated approximately 4,000 roles (5 per cent) in 2025, explicitly linking the cuts to the deployment of its Agentforce AI service, which the company claims reduces the need for human customer service agents.
- Intuit: Retrenched 1,800 staff (10 per cent) in mid-2024. CEO Sasan Goodarzi insisted the move was not about cost-cutting but about reallocating resources to AI, despite 1,050 of those cuts being attributed to underperformance.
- Dell Technologies: Reduced its headcount by roughly 12,000 between 2024 and 2025. Internal communications cited a reorganisation of go-to-market teams to focus on AI, while analysts noted a 150-basis-point drop in gross margins due to the high cost of AI server infrastructure.
- IBM: Has systematically replaced roughly 8,000 back-office roles with its AskHR AI platform since 2023, framing it as a workforce rebalancing rather than a traditional layoff.
The analysis suggests three critical reasons why the AI is taking jobs narrative is, in many cases, a self-serving fabrication.
The CapEx-OpEx Shell Game
AI is extraordinarily expensive. Microsoft has committed over $80 billion to AI infrastructure, and Cisco has allocated billions to AI startups while simultaneously cutting 7 per cent of its staff to lower costs. These vendors are not cutting people because AI is doing their jobs; they are cutting people to harvest the payroll (OpEx) required to pay for the massive hardware and energy costs (CapEx) that AI requires.
The Performance Mask
Vendors like Intuit and Salesforce have used AI-driven restructuring as a convenient PR blanket to exit thousands of employees. By framing these as AI pivots, leadership avoids the negative market sentiment usually associated with business as usual layoffs or admitting to a lack of fiscal discipline. Analysis shows that organisational bloat rather than AI is the underpinning issue in most cases.
The Hiring Spree Myth
A recurring pattern is the vendor’s claim that they will rehire an equal number of staff in AI-specific roles. In practice, this is often a tactical replacement of high-cost, experienced veterans with fewer, less experienced, lower-cost people. For the enterprise customer, this leads to a hollowing out of the vendor’s support and R&D divisions, leaving users with automated loop support that lacks the nuance to solve complex, real-world architecture problems. IBRS is already seeing this emerge, as clients report that SaaS release cycles are resulting in lower-quality, less consistent UX in solutions, and drastically poorer problem resolution. It is not that AI is ‘bad’, but rather it is not being driven professionally.
Who’s Impacted?
- HR Leads and Strategy Teams: Must navigate the internal cultural damage caused by vendor-fuelled AI fear. They must explain to their own executives that vendor layoffs are often a result of poor fiscal management, not a benchmark for their own AI readiness, and be far more pragmatic about the coming, gradual tide of employment change.
- The Board: Needs to question the long-term reliability of partners who are undergoing repeated rebalancing. They must look beyond the self-serving ‘AI made me do it, so you should too’ rhetoric regarding employment and look at the strategic and investment errors the vendors have made that left them with sharp declines in margin. They also need to expect ICT costs to rise sharply as these very same vendors restructure fees.
- CIOs/CTOs: Need to question the long-term reliability of partners who are undergoing repeated rebalancing. If a vendor is gutting its R&D and support to pay for GPU clusters, the quality of the product roadmap is at high risk. Must prepare for a decline in human-in-the-loop support. As vendors like Salesforce push Agentic AI as a replacement for staff, customers may find themselves paying premium prices for automated service levels that previously included human expertise.
Next Steps
- Demand Transparency: In your next QBR (quarterly business review) with major vendors, ask for a headcount breakdown in the specific support and engineering teams that service your account.
- Decouple AI from Efficiency: Reject vendor pitches that claim their tools allow you to reduce headcount. Demand evidence of service and role evolution rather than role elimination.
- Review ESG Disclosures: For organisations with strong social governance (ESG) commitments, evaluate whether your primary software partners are acting ethically in their workforce transitions or merely using AI as a mask to recover from systemic governance and strategy failures.
