Meta Taps Former Salesforce AI Chief To Lead Enterprise AI Push

Meta Platforms Inc. (NASDAQ: META) has appointed Clara Shih, former Salesforce Inc (NYSE: CRM) artificial intelligence chief, to head its newly formed Business AI division, marking a significant expansion of the social media giant's AI strategy for enterprise customers.

What Happened: Shih, who announced her move via LinkedIn on Tuesday, will spearhead Meta's efforts to develop AI tools for businesses using Meta's social platforms. "Our vision is to make cutting-edge AI accessible to every business, empowering all to find success and own their future in the AI era," Shih stated.

The appointment comes as Meta accelerates its unique open-source AI approach, differentiating itself from competitors like Microsoft Corp. (NASDAQ: MSFT) -backed OpenAI and Alphabet Inc.'s (NASDAQ: GOOG) (NASDAQ: GOOGL) Google.

Rather than monetizing AI through direct subscriptions, Meta aims to enhance its existing advertising and social media platforms using its Llama language models.

Why It Matters: Meta's AI push has gained momentum under CEO Mark Zuckerberg's leadership. During the company's third-quarter earnings call, Zuckerberg revealed plans for Llama 4, announcing an unprecedented computing infrastructure utilizing over 100,000 Nvidia Corp. (NASDAQ: NVDA) H100 GPUs. The next-generation model is scheduled for release in early 2025.

The company has already begun integrating AI-generated content into its platforms, including AI-created photo carousels on Facebook and AI chatbots on Instagram.

Bank of America Securities recently labeled Meta an "AI Story," citing the growing adoption of Llama and Meta AI. The firm projects AI-driven advertising improvements to materialize by 2025.

The move follows Meta's strong third-quarter performance, where the company exceeded analyst expectations while signaling increased capital investment in AI infrastructure. Zuckerberg has touted the company's $405 billion Llama 3.1 model as having superior cost performance compared to competing closed models.