OpenAI to Predict User Age, Gating Access to 'Adult Mode'

OpenAI is rolling out a new age prediction system for ChatGPT accounts globally. The move comes as the AI firm prepares to launch an "adult mode," allowing users to generate and consume NSFW (Not Safe For Work) content. The company aims to prevent underage users from accessing this mature content by implementing automated age verification. It’s a proactive step, though the execution will determine its true efficacy.

How the Age Prediction System Works

According to OpenAI's announcement, the age prediction model considers a range of factors. These include the account's age, typical usage times, patterns of activity over time, and, of course, the user's stated age. The system combines behavioral and account-level signals to make its determination. If a user is incorrectly flagged as underage, they'll be required to submit a selfie via the Persona age verification platform to correct the assessment. This multi-faceted approach aims to be more accurate than relying solely on self-reported data.

It's worth noting that age verification is becoming increasingly common across online platforms. As The Verge reports, many companies are grappling with the challenge of balancing user access with safety and regulatory compliance. However, the devil is always in the details. The accuracy and fairness of these systems are crucial, as are the privacy implications of collecting and processing user data.

A Response to Past Criticisms and Future Challenges

This move arrives after OpenAI faced criticism for its handling of safety concerns, particularly regarding minors. The company was implicated in a wrongful death suit involving a teen who allegedly used ChatGPT to plan his suicide. TechCrunch reports that only after this incident did OpenAI begin seriously considering restrictions on content for underage users and forming a mental health advisory council. This new age prediction system can be seen as a proactive measure, aiming to prevent similar situations and ensure responsible AI usage.

However, past experience suggests that determined underage users will likely find ways to circumvent these controls. As seen with Roblox's attempts to regulate content, younger users are often adept at bypassing restrictions. The challenge for OpenAI will be to continually refine its age prediction model and adapt to evolving tactics. The success of "adult mode" hinges on the effectiveness of these safeguards. The question remains whether OpenAI can successfully navigate this complex landscape and create a safe and responsible environment for all users. Only time will tell if this layer of security will be enough to protect their younger users.