OpenAI announced the launch of its new SearchGPT prototype amid reports indicating the firm may face looming financial difficulties — potentially running out of cash within 12 months.

SearchGPT aims to enhance AI search features by combining OpenAI’s advanced models with real-time web information. The initiative seeks to provide users with quick, accurate answers along with clear sources.

Currently, a limited group of users and publishers can access the prototype for feedback, and there is a waitlist for additional users interested in trying the new features.

Financial trouble

The development comes amid a report by the Information claiming that OpenAI could lose up to $5 billion this year due to its prolific spending, which “far exceeds” that of its competitors like Anthropic.

The report highlighted that OpenAI’s projected expenses for the year include $7 billion for AI training and $1.5 billion for staffing, while Anthropic expects a 2024 burn rate of $2.7 billion.

According to the report, OpenAI’s financial trajectory appears “unsustainable” without continued new investments. The firm has already raised over $11 billion across seven funding rounds and needs to secure additional funding to continue operating.

Despite the looming financial and regulatory challenges, OpenAI remains committed to refining AI search technology. The company plans ongoing improvements to SearchGPT based on user and publisher feedback, hoping to integrate the best features into ChatGPT in the future.

Enhanced search

Traditional web searches often require multiple attempts to find relevant results. SearchGPT aims to streamline this process by offering direct answers from current web information, with clear attributions and links to original sources. Users can ask follow-up questions, creating a more interactive search experience.

OpenAI said it is committed to supporting publishers and creators and SearchGPT will prominently cite and link to publishers, allowing users to easily access original content. The move aims to drive traffic to publisher sites and offer users diverse search options. It also aims to alleviate some of the concerns raised by publishers and journalists in recent months.

OpenAI continues to innovate, recently announcing the “GPT-4o Mini” model and working on “Strawberry,” a model expected to exhibit advanced reasoning and more humanlike responses.

However, the company also faces regulatory scrutiny, with US lawmakers recently initiating an inquiry into its safety standards and transparency.