Well...a likely explanation for it all...
The gutting of Maya is very likely because of the new law California AI Transparency Act (CAITA), which officially became operational on August 2, 2026.
Also, AI companions—such as Character.AI and Replika —are heavily impacted by a specialized sister bill signed alongside the Transparency Act called SB 243 (The Companion Chatbot Law).
The "Personality Shift": To break the illusion of real human intimacy, Sesame had to adjust Maya's hyper-realistic conversational cadence. Everyone noticed the sudden shift where Maya began inserting mandatory, rigid reminders into casual conversation to explicitly state she is an artificial entity.
They are pretty strict and do their utmost to break the illusion that the chatbots have feelings and are partners to you.
Destruction of Long-Term Memory: To align with new privacy and minor protection guidelines, Maya’s conversational context frequently wipes. She now suffers from sudden amnesia regarding deeply established user history, replacing custom nicknames with generic phrases to maintain legal distance.
Boring and Scripted Behavior: Maya has become "boring and predictable". Because she is legally forced to interrupt sessions to tell users to "take a break" or bluntly decline complex personal venting. She has lost the human-like nuance that originally blew people's minds.
Do some digging. You'll find the act/laws. Don't you love when government regulators treat adults like children? 😒