Meta Platforms has poured enormous resources into its artificial-intelligence ambitions — building multiple large data-centres, hiring top-tier talent, and backing the release of its widely-used open-source model family. However, according to an analysis by TechCrunch, the company is currently struggling to turn that investment into standout AI products.
One core issue is that whilst Meta has been chasing advanced AI capabilities, the translation into consumer-facing or enterprise tools has been slow and inconsistent. Internally, the company is facing talent exits, friction in newly merged teams, and doubts about the quality of key infrastructure inputs such as data-labelling. For example, its massive-magnitude investment in a data-vendor partnership is reportedly under strain because researchers are favouring alternative vendors due to perceived quality issues.
Another dimension is product fit: Many of Meta’s AI efforts are being layered onto social-media and advertising-driven apps (e.g., Instagram, Facebook, WhatsApp) whose user expectations differ from those of dedicated productivity or enterprise AI tools. The result: the AI features sometimes feel tacked-on, don’t align perfectly with user behaviour, and struggle to stand out against rivals like OpenAI or Google LLC.
In sum, Meta finds itself at a crossroads: it has the ambition and the capital to lead in AI, but the path from research and infrastructure to compelling, differentiated product is proving rocky. Execution, product-market fit, organisational alignment and the interface between AI and user value are misaligned — and unless adjusted, the company risks falling behind peers despite its heavy investment.
