AI giants race to scoop up elusive real-world data

Amid an intense battle for supremacy, artificial-intelligence companies are forging alliances across industries and regions to help gather real-world data that can’t be scraped from the internet.

Over the past two months, OpenAI has tied up with e-commerce majors ShopeeiShopeeSingapore-based Shopee, a subsidiary of Sea Ltd., is a major e-commerce platform that operates across Southeast Asia.READ MORE and Shopify, while Google and Perplexity have doled out free access to their advanced AI tools to some users in India. Experts believe these moves will help the companies access structured consumer queries, product behaviors, and transactional data — training signals that are often unavailable via public data alone.

“These partnerships will provide them with diverse data sets that will help them to train their AI models better and generate more accurate outputs,” Sameer Patil, director of the Centre for Security, Strategy & Technology at global think tank Observer Research Foundation, told Rest of World. “It will also help them to innovate new ways to apply AI models in some particular sectors. This particularly applies to sectors where there is emphasis on hyper-customization and hyper-personalized offerings like fintech and health care.”

“Asia’s young, tech-savvy population and high mobile penetration make it one of the fastest-growing markets for AI adoption and innovation,” Oliver Jay, managing director of OpenAI International, said in a statement announcing the partnership with Sea. “By applying OpenAI’s capabilities in areas like commerce, we look forward to working with Sea as they equip more businesses with smarter tools to manage operations and reach customers.”

Google, Perplexity, and OpenAI did not respond to specific queries about why they were entering global partnerships or offering their AI tools for free to some users. According to their privacy policies, all three companies collect user data for reasons like improving experience, creating and managing accounts, and training their models. Users can opt out of sharing their data to train models.

China has already demonstrated how access to industry-specific data can provide a competitive edge.

Earlier this month, Rest of World reported that China’s AI drug discovery companies have landed multibillion-dollar deals with pharmaceutical companies like AstraZeneca, Pfizer, and Sanofi. The success is partly driven by the fact that Chinese companies have access to a massive contextual data set thanks to the country’s national health insurance system, which covers over 600 million people. Scott Moore, director of China programs at the University of Pennsylvania, told Rest of World this is a “structural advantage,” because the large patient pool can be used as a vast training set for training AI models.

Google and Perplexity are seeing early success with their free launches in India.

In July, the month when Perplexity announced its deal with Bharti Airtel, the company clocked monthly downloads of 6.69 million, compared to just 790,000 in June, according to data from market intelligence firm Sensor Tower.

“I downloaded it [Perplexity] last month because it is free with Airtel now. I use ChatGPT more regularly,” P. Sahay, a data scientist who works with a consulting firm in Bengaluru, told Rest of World. “I am not too worried about my data because the kind of things I use AI for are not very sensitive. I use it for framing emails or messages, code debugging, to get access to off-the-shelf codes, market research, etc. None of that has confidential information as such.” Sahay said he was not aware that AI chatbots offered the option to turn off data collection for training.

Even as users lap up the benefits offered by AI giants, experts are concerned about issues such as data sovereignty and a trend where emerging markets become feeders into global AI systems without fair returns.

Several developing countries in Africa and Asia have been pushing back on Big Tech’s hold on global data. Countries such as Nigeria, India, South Africa, and Vietnam have been asking companies like Google, Microsoft, and Amazon to store their citizens’ information locally.

“This calls for robust guardrails to protect privacy and ensure fairness,” Patil said. “Participating companies will have to ensure that the data sets are non-personalized and anonymized, as well as robust oversight frameworks are put in place for avoiding biases in data harvesting.”

Even as the jury is out on how far these freebies and partnerships can take AI companies, the supremacy battle took a fresh turn on August 12 when Perplexity offered to purchase Google’s Chrome browser for $34.5 billion. The offer price is nearly double Perplexity’s valuation of $18 billion, The Wall Street Journal reported.