Orkut Could Return: Founder Collects 270,000 Signups

Orkut Büyükkökten has asked users to visit Orkut.com and add their names in support of bringing Orkut back. The request came in an X post on Thursday, according to TechCrunch. More than 270,000 people had signed up as of October 9, 2026, according to YourStory.

Orkut started in 2004 as a side project by Büyükkökten while he was at Google. It became very popular in India and Brazil before losing users to Facebook. Google closed Orkut in September 2014. At its peak, it had about 300 million active users.

Orkut.com now says, "You deserve better from social media." The site says adding your name shows support for a new social network built around "authentic people, real connection, and genuine belonging." Visitors are asked to add their name to join the movement and send a supporter card to one other person to join them.

In a letter on Orkut.com, Büyükkökten said "what people miss most about orkut.com is the communities." He pointed to "communities of people who repaired the same motorcycle, rode the same bus route, and hated Mondays" as examples. In the same letter, he criticized the growing role of algorithms, the automated systems that decide what posts you see, and AI-generated content on social media. He called for an internet built around real people "with a human on the other side of every message, not focused on algorithms, eyeballs and time-on-site optimization." He has also flagged bots, AI-generated content and deepfakes in connection with the revival effort.

This is not his first try at a successor. Büyükkökten started a social network called Hello in 2016 to connect people through shared interests. Hello closed in 2020 after more than 2 million downloads worldwide. In 2022, he said on Orkut.com that he was building something new. In 2024, he told WIRED he was working on a social network that would use AI and machine learning, software that learns patterns from data, to bring people together and build communities. He has not said whether a new version of Orkut is already being built or whether it will ever launch. No launch date had been confirmed as of October 9, 2026.

The broader context here will be familiar to anyone who has run an online community. The original Orkut was organized around the social graph, the map of who knows whom, and user-created communities, where membership, moderation and threaded discussion decided what stood out. Modern feeds reverse that setup, with recommender systems, software that ranks posts to increase engagement and time on site, filling in social signals afterward.

In my view, that reversal explains why the revival pitch focuses on communities rather than content discovery.

For builders, there is a practical problem worth flagging. Nostalgia alone does not start a network. Interest-based communities face a cold-start problem, meaning an empty group gives new members no reason to stay. Safety tools must handle Sybil accounts, large numbers of fake identities controlled together, plus automated content and synthetic media at launch, not later. Identity checks, rate limits and proof-of-personhood tests all make signing up harder, which slows growth. That history matters. Hello tested the shared-interests idea and did not reach lasting scale.

Looking ahead, the opening is still easy to see. Small groups with clear membership can support stronger moderation rules and lower inference costs, the computing cost of running AI models, than open-ended recommendation. They can also make proof of origin and reputation carry across groups. Whether 270,000 email signups turn into daily members who start and moderate those spaces is the test. A list shows interest. It does not create belonging.