Apple’s AI shopping wishlist: If not Perplexity then what… 

Apple

Apple’s search for a significant AI acquisition has become more urgent, as internal talent departures continue and the company’s own AI product rollout lags behind rivals.

Speculation over a potential purchase of Perplexity has diminished since the company was reportedly found bypassing content restrictions. At the same time, Apple’s Foundation Models team has seen a steady outflow of engineers, intensifying pressure to secure both talent and technology.

Reports, though unconfirmed, suggest Apple made over 30 AI-related acquisitions in 2023, surpassing the numbers of Google, Meta, Microsoft, and Amazon. Despite this activity, industry observers note that Apple’s AI progress has not matched expectations. Tools such as the AI-powered Writing Tools announced a year ago have seen limited updates, while Siri’s planned revamp has faced repeated delays and technical challenges.

Analysts point out that Apple may now need more than engineering talent. The focus is shifting towards acquiring a company with proven, product-ready AI systems, preferably backed by its own foundation models, according to a report by 9to5 Mac.

Several candidates are effectively ruled out. OpenAI and Anthropic are considered too large or strategically incompatible. Perplexity could contribute to Siri’s capabilities, but its valuation of around $18 billion and broader focus on e-commerce, sports, finance, and search integration suggest it may not be the best cultural or strategic fit for Apple.

One alternative is Mistral, the French AI firm valued at $10 billion and known for its foundation models, open-source tools, consumer applications, and developer APIs covering OCR, speech, coding, and workflow automation. Mistral’s approach to privacy and responsible AI aligns closely with Apple’s public positioning.

Another possibility is Germany’s Aleph Alpha, which has raised close to $1 billion and specialises in enterprise and public-sector AI. While this focus differs from Apple’s consumer-oriented products, it would not be the first time the company has adapted technology from outside its core market, as was the case with Siri’s early development. However, retention could prove difficult if the team remains primarily motivated by non-consumer projects.

Beyond Europe, options thin further. The Allen Institute for AI (Ai2) in the US continues to advance open research and tools. China-based Kimi has produced the competitive K2 model, while Canada’s Cohere targets developer and enterprise solutions. Other emerging companies include Thinking Machines Lab, led by former OpenAI executive Mira Murati, valued at $12 billion, and Safe Superintelligence, co-founded by Ilya Sutskever, valued above $30 billion despite being in an early stage without a product.

With valuations for leading AI labs climbing rapidly, analysts suggest Apple’s acquisition window is closing. At present, Mistral appears to offer the best combination of technological capability, cultural alignment, and strategic value to strengthen Apple’s position in AI before the competitive gap widens further.