Mark Zuckerberg Drives Meta's AI-Driven Transformation with Stringent Engineering Guidelines

Meta Aims to Achieve AI-Native Status with Aggressive Adoption Targets

Meta, the parent company of Facebook, is pushing forward with its vision of becoming an "AI-native" company. A recently leaked internal document reveals the company's targets for AI adoption across various departments.

Key Targets:

• 65% of engineers in Meta's creation org are expected to write 75% of their committed code using AI tools by the first half of 2026.
• 50% to 80% of code in the Scalable Machine Learning organisation is expected to be AI-assisted by February 2026.
• 80% of mid- to senior-level engineers in the central products division are expected to adopt general AI tools, such as DevMate, Metamate, and Google's Gemini, by Q4 2025.
• 55% of all software engineer code changes are expected to be "Agent-Assisted" by Q4 2025.

Read also: Treasury Yields Experience Largest Increase in Two Weeks Following Release of Labor Market Data

Company-Wide Goals:

• The central products division aims to have 55% of code changes from software engineers be "Agent-Assisted" by Q4 2025.
• The Scalable ML Team aims to have 50-80% of code written being 'AI-assisted' by 2026.

Performance Reviews:

While it is unclear if the goals listed in the document are tied to performance reviews, a Meta spokesperson stated that the company's performance programme is focused on rewarding impact from AI tools, rather than just usage.

Read also: US-Iran Tensions Spark Uptick in Oil Prices Amid Global Market Decline