In late March, Oracle terminated approximately 20,000 to 30,000 employees across its organization via email, followed days later by a severance offer that would spark immediate controversy among affected workers. The package itself appeared conventional on the surface: four weeks of base salary plus one additional week per year of tenure, capped at 26 weeks, along with one month of COBRA insurance coverage. But this apparent reasonableness masked a far more punitive reality—Oracle declined to accelerate any unvested restricted stock units (RSUs), meaning thousands of employees forfeited stock grants that had been positioned as compensation and retention incentives. For some long-tenured staff, this amounted to hundreds of thousands or even over a million dollars in equity that would never vest. The company also aggressively interpreted WARN Act requirements, classifying remote workers in non-protective states as ineligible for the two-month notice period, further narrowing the severance floor.
Oracle's actions arrived at a specific inflection point in the AI boom cycle. After a golden age of aggressive hiring and recruitment bonuses in 2023-2024, the industry has shifted toward aggressive cost-cutting justified by the need to achieve "AI efficiency." Unlike the slow-motion recessions of 2022-2023, where layoffs spread across the year, the Oracle cuts represented a singular, coordinated purge timed to deliver a quarterly benefit to investors. This reflected a hardening consensus among enterprise software leaders that they had over-hired and needed to recalibrate headcount quickly. But Oracle's approach also signaled something else: a willingness to be harder on departing employees than the market norm, banking on the assumption that former workers would lack the leverage to push back—a calculation that proved partially wrong when at least 90 employees organized to petition for better terms.
The severance dispute cuts to the heart of how tech compensation has been restructured over the past decade. Base salary at companies like Oracle has stagnated or declined as a percentage of total package, while equity grants—particularly for engineering and product talent—have grown to represent 50, 60, or even 70 percent of annual compensation. This shift was marketed to employees as an alignment mechanism, tying their fate to company success. What it actually created was a system where the company could shed headcount while preserving cash, simply by forfeiting the paper wealth of RSUs that had not yet vested. An employee with four months of unvested grants loses those grants entirely; an employee six months away from vesting loses thousands in real future compensation. Oracle's decision to decline acceleration treated these unvested units as discretionary generosity rather than deferred wages—a legal distinction that carries profound practical consequences for workers.
The impact fell disproportionately on Oracle's most experienced workforce. Staff who had accumulated five, ten, or fifteen years at the company faced the steepest losses, since longer tenure correlated with larger equity grants and more unvested tranches on the termination date. Junior employees, who had smaller equity grants and shorter vesting schedules, absorbed proportionally smaller hits. This created a perverse outcome: the most expensive employees to retain—those with deep context and institutional knowledge—became the most expendable, since shedding them reduced the severance cash outflow. The remote worker WARN Act carve-out further narrowed protections for geographically distributed staff, another cost-minimization technique that corporate counsel had presumably vetted and approved.
The competitive angle here matters enormously for the broader labor market. Meta, by contrast, offered 16 weeks of base salary plus two weeks per year of service, along with 18 months of COBRA, according to publicly disclosed information. Amazon, Google, and other mega-cap tech firms have consistently accelerated RSU vesting during reductions in force, treating unvested compensation as a moral and reputational obligation. Oracle's refusal to match this norm signals a deliberate choice to be the hardest negotiator in the room—a signal aimed at both departing employees and Wall Street investors. For candidates evaluating job offers, it also introduces a new risk factor: the company's willingness to shed equity acceleration during downturns. That reputational cost may not be fully priced into Oracle's calculation yet.
The open question now is whether Oracle's approach becomes a template or remains an outlier. If other enterprise software vendors adopt similar severance structures—declining to accelerate RSUs, aggressively interpreting WARN Act scope—the practice could begin normalizing across the industry. If the organized employees' petition gains traction or escalates to legal scrutiny, Oracle may face pressure to recalibrate. The deeper implication is how the industry resolves the equity acceleration question during the next wave of efficiency-driven layoffs, which now seems inevitable as AI capex pressures mount and hiring discipline tightens. What was sold to employees as retention and alignment may ultimately be remembered as a mechanism to shift downside risk onto the workforce.
This article was originally published on TechCrunch AI. Read the full piece at the source.
Read full article on TechCrunch AI →DeepTrendLab curates AI news from 50+ sources. All original content and rights belong to TechCrunch AI. DeepTrendLab's analysis is independently written and does not represent the views of the original publisher.
