Perplexity's CEO says it's still aiming for a 2028 IPO, regardless of how OpenAI and Anthropic fare

Perplexity, the AI-driven search and answer engine startup, is resolute in its plans for a 2028 initial public offering (IPO), despite the anticipated high-profile market entries from competitors such as OpenAI, Anthropic, and SpaceX. According to CEO Aravind Srinivas, the company’s timeline remains unaffected by how these industry giants fare in the market. In a recent interview with CNBC, Srinivas emphasized that Perplexity's strategy is based on its own roadmap, stating, "Agnostic of these two companies, we were planning for something in 2028, so that still remains the case." His remarks come as the AI sector braces for a potentially bustling period of technology IPOs. Anthropic, known for its Claude AI, has recently filed for an IPO, while OpenAI has also taken steps towards a public offering. Additionally, SpaceX’s upcoming market debut is expected to gauge investor interest in large-scale tech offerings. Collectively, these events could significantly influence how the public markets perceive the value of emerging AI and tech leaders. Srinivas acknowledged that the success or failure of these IPOs will have broader implications for the industry, but reiterated that they would not sway Perplexity's own ambitions. "I certainly think there will be ripple effects if they don't go well, like there is no sugar coating on that," he noted. He also defended the high valuations associated with OpenAI and Anthropic, asserting that both are at the forefront of AI innovation. However, he cautioned that investors will closely monitor whether these leading firms can sustain their rapid advancements. "If for six months you don't see a model capability advance from one of these two companies, then it's a problem for them," he remarked. As the conversation about AI valuations intensifies, companies are reassessing their AI expenditures. Srinivas pointed out that clients are becoming increasingly discerning regarding their choice of models. "If there is an open source model that gets the job done 90% of the time, I'd probably use that if it's 10 to 20 times cheaper than the frontier model," he explained.