Meta AI Avocado Delayed as 20% Workforce Cuts Loom

Meta AI Avocado Faces Major Setbacks

Meta AI is facing one of the most dramatic challenges in its corporate history with the Meta AI Avocado delay. The social media giant has announced plans to cut approximately 20% of its workforce—potentially affecting over 15,000 employees—while simultaneously pouring $135 billion into artificial intelligence research and development this year alone. This represents a doubling of AI investment compared to 2025, according to reports from The New York Times and Reuters. The strategic pivot comes as Meta races to catch up with AI rivals including OpenAI, Google, and Anthropic, but faces significant internal challenges in delivering competitive AI products.

The timing of these announcements has raised eyebrows across Silicon Valley. Meta's stock dropped nearly 4% on Friday following reports of the Meta AI Avocado delay, adding to investor concerns about the company's ability to compete in the rapidly evolving AI landscape. Despite having assembled an elite internal unit called TBD Lab specifically to build breakthrough AI systems, Meta has yet to deliver a model that can match the capabilities of leading competitors. The company's aggressive hiring strategy, which has involved recruiting top AI talent with compensation packages that are "tens to hundreds of times more" than regular employee salaries, has not yet translated into competitive products.

Meta AI Model Falls Short of Expectations

Meta AI's upcoming foundational model, code-named "Avocado," has been postponed from its original March 2026 launch target to at least May. The delay comes after internal testing revealed that the model failed to match the performance of leading AI systems from OpenAI, Anthropic, and Google. According to The New York Times, internal tests showed Avocado falling short in key areas including reasoning, coding, and writing capabilities. While Avocado outperformed Meta's previous AI model and even surpassed Google's Gemini 2.5 from March 2025, it lagged significantly behind Gemini 3.0 from November 2025.

The underperformance has led Meta executives to consider temporarily licensing Google's Gemini models to power some of Meta's AI products, according to reports from Benzinga and other news outlets. This potential partnership would represent a significant shift in Meta's strategy, as the company had originally planned to compete directly with Google's AI offerings. The TBD Lab, led by Meta's AI chief, has been working on Avocado for months, but the team has only released one public product so far: Vibes, an AI video app that failed to gain significant traction in the market.

Despite these setbacks, Meta remains committed to its AI ambitions. The company is already planning an even larger model codenamed "Watermelon" that would follow Avocado, according to Chosun Ilbo. Meta has also acquired the AI social media platform Moltbook to strengthen its position in the AI assistant market. The company's AI spokesperson stated that the next model would "show the rapid trajectory we're on" and that Meta would "steadily push the frontier over the course of the year as we continue to release new models."

Workforce Reductions Amid Record AI Spending

The decision to cut 20% of Meta's workforce while simultaneously increasing AI investments reflects the company's belief that AI-driven efficiency can replace human labor across many functions. According to Investor's Business Daily, the layoffs could affect approximately 15,800 of Meta's nearly 79,000 employees. This would represent the largest layoff in Meta's history and comes just two years after the company cut 11,000 jobs in 2023. The move aligns with a broader industry trend toward AI-assisted operations, where companies automate routine tasks previously performed by human workers.

Meta's $135 billion AI investment for 2026 is part of a massive infrastructure buildout that includes plans to invest $600 billion in data center construction by 2028. This represents one of the largest capital expenditure commitments in corporate history. The company has been acquiring AI talent aggressively, spending hundreds of billions of Korean won to recruit key researchers from competing companies. However, a Meta AI researcher recently stated that "it's unreasonable for people receiving tens to hundreds of times more compensation than regular employees to fail to deliver results," highlighting the pressure on the AI team to deliver competitive products.

The combination of massive AI spending and workforce reductions has created a challenging situation for Meta. Investors have shown concern, with the stock declining 23% from its recent highs as questions mount about the company's AI strategy. While Meta's advertising business continues to generate significant revenue from its 3.35 billion daily active users across Facebook, Instagram, and WhatsApp, the company faces an uphill battle to gain market share in the AI model market. Analysts at Forbes noted that Meta is "losing badly" in the competition for developers and enterprises paying for API access to AI models.

The company's struggles with AI development extend beyond Avocado. Before the Avocado project, Meta scrapped the largest version of its Llama 4 model, called Behemoth, which had been scheduled for release last summer after facing criticism for delivering misleading benchmark results. This pattern of setbacks suggests that Meta AI division faces fundamental challenges in competing with more established AI research teams at OpenAI, Google, and Anthropic. The Meta AI Avocado delay represents just the latest chapter in the company's ongoing AI struggles.