“AI is dear. Let’s be sincere about that,” Anand says.
Progress vs. Security
In October 2024, the mom of a teen who died by suicide filed a wrongful dying go well with towards Character Applied sciences, its founders, Google, and Alphabet, alleging the corporate focused her son with “anthropomorphic, hypersexualized, and frighteningly lifelike experiences, whereas programming [the chatbot] to misrepresent itself as an actual individual, a licensed psychotherapist, and an grownup lover.” On the time, a Character.AI spokesperson instructed CNBC that the corporate was “heartbroken by the tragic loss” and took “the security of our customers very severely.”
The tragic incident put Character.AI underneath intense scrutiny. Earlier this 12 months, US senators Alex Padilla and Peter Welch wrote a letter to a number of AI companionship platforms, together with Character.AI, highlighting considerations about “the psychological well being and security dangers posed to younger customers” of the platforms.
“The staff has been taking this very responsibly for nearly a 12 months now,” Anand tells me. “AI is stochastic, it is type of onerous to all the time perceive what’s coming. So it isn’t a one time funding.”
That’s critically essential as a result of Character.AI is rising. The startup has 20 million month-to-month lively customers who spend, on common, 75 minutes a day chatting with a bot (a “character” in Character.AI parlance). The corporate’s consumer base is 55 p.c feminine. Greater than 50 p.c of its customers are Gen Z or Gen Alpha. With that progress comes actual threat—what’s Anand doing to maintain his customers secure?
“[In] the final six months, we have invested a disproportionate quantity of assets in having the ability to serve underneath 18 in a different way than over 18, which was not the case final 12 months,” Anand says. “I can not say, ‘Oh, I can slap an 18+ label on my app and say use it for NSFW.’ You find yourself creating a really completely different app and a special small-scale platform.”
Greater than 10 of the corporate’s 70 workers work full-time on belief and security, Anand tells me. They’re liable for constructing safeguards like age verification, separate fashions for customers underneath 18, and new options similar to parental insights, which permit mother and father to see how their teenagers are utilizing the app.
The under-18 mannequin launched final December. It contains “a narrower set of searchable Characters on the platform,” in keeping with firm spokesperson Kathryn Kelly. “Filters have been utilized to this set to take away Characters associated to delicate or mature subjects.”
However Anand says AI security will take extra than simply technical tweaks. “Making this platform secure is a partnership between regulators, us, and oldsters,” Anand says. That’s what makes watching his daughter chat with a Character so essential. “This has to remain secure for her.”
Past Companionship
The AI companionship market is booming. Shoppers worldwide spent $68 million on AI companionship within the first half of this 12 months, a 200 p.c enhance from final 12 months, in keeping with an estimate cited by CNBC. AI startups are gunning for a slice of the market: xAI launched a creepy, pornified companion in July, and even Microsoft payments its Copilot chatbot as an AI companion.
So how does Character.AI stand out in a crowded market? It takes itself out of it completely.
