Something new is happening in tech. In addition to building their own AI models, companies are now racing to buy startups.

Apple and Meta, among others, are looking to acquire promising young AI firms in a bid to stay ahead in what many are calling an “AI gold rush”. Startups like Perplexity AI, Safe Superintelligence (SSI), and Thinking Machines Lab have suddenly become hot property, each now worth billions.

Apple appears to be struggling with its AI strategy, a fact highlighted at the WWDC 2025 event, where Siri received no new updates. Moreover, instead of showcasing in-house developments, the company has been relying heavily on its partnership with OpenAI to stay relevant in the AI race.

Apple is looking to acquire Perplexity AI, a startup founded by Aravind Srinivas, an IIT Madras graduate. Known for its AI-powered search engine and chatbot platform, Perplexity is currently valued at around $14 billion (approximately ₹1.21 lakh crore).

According to the Bloomberg report, the tech giant is exploring an AI search engine for Safari as it prepares for the possible end of its long-standing deal with Google. The move would also counter Samsung’s reported plans to integrate Perplexity AI into its Galaxy S26 series as the default assistant by 2026. 

However, Perplexity’s leadership has downplayed the talks, stating they have “no knowledge of any current or future M&A discussions”. This suggests that the deal remains purely speculative at this stage.

Notably, Meta also considered acquiring Perplexity AI and Thinking Machine Labs before investing in Scale AI. These acquisitions are important for big tech companies in many ways, as they want to acquire their technology and talent.

Talent Acquisition 

AI talent is hard to find, and startups often bring together some of the best minds in the field. Acquiring one helps companies build strong teams without starting from scratch. Notably, in a recent podcast, OpenAI CEO Sam Altman said that Meta is offering extremely high compensation packages to OpenAI employees, including signing bonuses of up to $100 million.

Similarly, Anthropic is also engaged in a talent war with OpenAI and Google DeepMind.

“Whereas one year ago the narrative was around pre-training compute requirements driving consolidation, today the narrative is all about talent advantages being critical in a world of increasing compute abundance,” David Cahn, partner at Sequoia Capital, wrote in an article. 

Cahn added that Meta’s decision to acquire a 49% stake in Scale and bring in CEO Alexandr Wang to lead their new “founder mode” AI lab is an even clearer move in this direction. According to him, breakthroughs in reinforcement learning or elsewhere will require talent. 

“With their obsessive focus on talent, the AI labs are increasingly looking like sports teams: They are each backed by a mega-rich tech company or individual. Star players can command pay packages in the tens of millions, hundreds of millions, or, for the most outlier talent, seemingly even billions of dollars,” his article stated. 

However, he added that, unlike sports teams where players often have long-term contracts, AI employment agreements are short-term and liquid. This means anyone can be poached at any time. 

Big Tech’s Struggles and Desperation

Both Apple and Meta have been facing setbacks in their AI strategies. Meta’s latest model, Llama 4, didn’t receive positive feedback from the developer community

and was launched with a few complaints and controversies surrounding the benchmarks. 

While the company denied any wrongdoing, its public image was affected. The company also recently saw notable talent departures, with some joining French AI firm Mistral.

Moreover, Meta reportedly made an acquisition offer for Safe Superintelligence (SSI), the AI startup co-founded by former OpenAI chief scientist Ilya Sutskever, which had reportedly been valued at around $32 billion during its fundraising efforts. However, the offer was declined. 

Nonetheless, Meta is trying to hire SSI co-founder Daniel Gross and former GitHub CEO Nat Friedman to strengthen its new AI division, which is headed by Wang. 

Meta has reorganised its AI efforts into two core divisions. The one headed by Connor Hayes focuses on integrating AI features into Facebook, Instagram, WhatsApp, and the Meta AI assistant. The other, which is called AGI Foundations, is led by Ahmad Al-Dahle and Amir Frenkel and is dedicated to core AI research, including work on the Llama models. 

Although the FAIR research lab, established by Meta chief AI scientist Yann LeCun and previously overseen by Joelle Pineau until she left in May, continues to operate independently, some of its teams have been moved into the new structure.

Zuckerberg has never shied away from bold acquisitions. One of the most notable came in April 2012, when Facebook (now Meta) purchased Instagram for roughly $1 billion in cash and stock. The app, which had launched just two years earlier, had already attracted millions of users with its minimalist interface and iconic photo filters.

On the other hand, historically, Apple’s largest acquisition was Beats in 2014 for $3 billion. The Perplexity deal, if it happens, would be the biggest in Apple’s history, reflecting the company’s urgency to catch up on AI.

The company was recently sued by shareholders in a class action, alleging it misrepresented how long it would take to upgrade Siri with advanced AI, a misstep they say dragged down iPhone sales and the stock price.

Valuations Without Products?

Notably, Thinking Machines Lab and SSI have no products to show yet. Their high valuations stem primarily from the reputation and talent of their founders. The former reportedly closed a $2 billion seed round, valuing the six-month-old startup at $10 billion, according to the Financial Times.

However, when Murati left OpenAI and founded Thinking Machine Labs, she brought around 20 OpenAI employees with her and later expanded the team to 60 by hiring from OpenAI and other AI labs.

At the same time, SSI intends to scale AI differently, giving as much importance to safety research as to performance.

Perplexity, while not building its own state-of-the-art models, has instead created a sophisticated orchestration of multiple LLMs, allowing users to choose based on their specific needs. OpenAI co-founder Andrej Karpathy even cited it as a prime example of multi-model architecture paired with human-in-the-loop verification

The company has been actively rolling out new products. The latest is Comet, its agentic browser. It reimagines web browsing through context-aware intelligence. 

Comet personalises responses based on a user’s browsing history and open tabs, all of which are stored locally, and not used for model training. It is now available on Windows with features like agentic search, shopping cart discounts, and email assistance. 

Besides, Perplexity offers finance-focused features, including a stock screener for filtering investments, earnings call transcript generation for publicly traded companies, comprehensive company profile creation, and a live dashboard displaying real-time stock data, market summaries, and upcoming earnings announcements. 

Last year, reports claimed that Meta is also planning to build its own search engine. This acquisition might be the answer to that. 

Startups now possess the talent and momentum that large tech companies require. Building from the ground up takes too long, and time is a luxury few can afford. Acquisitions are no longer optional; they are urgent. Perhaps, in the race for AI leadership, owning the correct startup can make all the difference.

The post Why Apple and Meta Are Ready to Throw Billions on This IIT Madras Graduate’s AI Startup appeared first on Analytics India Magazine.