OpenAI’s ChatGPT platform has sparked a viral social media sensation, with Kenyans re-feeding their personal data, including selfies, weeks after Sam Altman’s Worldcoin deleted all biometric data collected from Kenyan citizens.
To generate a caricature, ChatGPT relies on content actively provided by the user, such as uploaded images, prompts, and descriptive information. Data Commissioner Immaculate Kassait warned Kenyan citizens that this practice shares biometric data freely, training AI models like OpenAI’s without user compensation.
Kenyans freely give their biometrics to Altman’s bot
This trend follows Kenya’s battle to delete its data from Worldcoin. Worldcoin is a crypto-backed digital identity project co-founded by OpenAI CEO Sam Altman. It is designed to verify that users are real humans by scanning their irises with specialised hardware in exchange for a digital ID and crypto tokens.
Worldcoin’s operations in Kenya began in 2023 with the deployment of its signature “orb” devices, spherical scanners that captured iris and facial scans from participants. In exchange, users were offered 25 free Worldcoin tokens (Ksh 8,256 at the time) at the time. This incentive drew large crowds but also deep concern from privacy advocates and regulators.
Kenya’s regulatory pushback was swift. In August 2023, the government suspended the project amid fears the data could be misused or transferred outside the country without sufficient safeguards.
In May last year, the High Court ruled that the company behind Worldcoin violated the Data Protection Act of 2019 by collecting sensitive personal data without proper consent and without conducting a mandatory Data Protection Impact Assessment (DPIA).
Last month, Julisha.co.ke reported that the Office of the Data Protection Commissioner (ODPC) confirmed that all iris scans and other biometric identifiers gathered during Worldcoin’s 2023 enrolment exercise had been erased from the project’s systems.
However, with the new AI caricatures is obtaining the very data that was deleted. During the process of creating it, ChatGPT requests a clear selfie, good lighting, a visible face, and no heavy filters.
“I need your actual face […] once I have it, I can exaggerate the right features, sharpen the attitude, and dial the realism just right,” the bot states.
To Ms Kassait, the AI caricatures trend is part of a phenomenon called surveillance capitalism. According to Harvard, it is a new form of capitalism that converts human behaviour into data for tracking, analysis, and monetisation.
“What you have just done is share your biometrics. In the future, somebody doing analytics can actually tell every single thing about you. You clicked, you didn’t ask what the purpose is,” she added. To that end, she encouraged users to carefully read the platform’s terms before sharing.
OpenAI expands Ad model as financial pressures mount
The caricature trend emerges as AI companies face financial pressures. OpenAI recently introduced advertising in ChatGPT responses and launched a cheaper $8 monthly Go tier to increase revenue.
OpenAI has approximately $13 billion in revenue and around $1.4 trillion in compute commitments. The AI company announced a string of partnerships in recent months. Among them, Nvidia said it would commit $100 billion to support OpenAI as it builds and deploys at least 10 gigawatts of Nvidia systems.
More recently, OpenAI announced a $10 billion deal with chipmaker Cerebras to deploy 750 megawatts of Cerebras’ AI chips. It also has other agreements with AMD and Broadcom.
Meanwhile, Anthropic and OpenAI are publicly clashing over their business strategies, especially the role of advertising in AI products. Anthropic paid for high-profile Super Bowl ads mocking OpenAI’s move to start showing ads in ChatGPT, using the tagline “Ads are coming to AI. But not to Claude.”
Anthropic has committed $50 billion to building data centers in the US, but it will also spend money buying computers from players like Microsoft and Google. According to Anthropic, unlike OpenAI, it’s doing more with less.
