Why Twitter’s New CEO Is Tackling Algorithmic Bias and What It Means for AI
Jack Dorsey stepped down as Twitter CEO, appointing CTO Parag Agrawal—an AI‑focused leader who aims to reduce algorithmic bias, advance open‑source social standards, and drive growth—while the company’s stock rose and broader industry trends in leadership were highlighted.
Jack Dorsey, Twitter’s founder and CEO, announced his resignation on Monday, naming CTO Parag Agrawal as his successor and a new board member.
Agrawal, who joined Twitter in 2011 and served as CTO since 2017, is a 37‑year‑old veteran who helped grow the user base and earned the company’s first Distinguished Engineer award.
During his tenure he has focused on reducing algorithmic bias in machine‑learning systems, such as the image‑cropping model that favored white faces, and he publicly pledged to make Twitter’s AI more transparent.
He also leads the Blue Sky initiative, an effort to fund engineers building open, decentralized standards for social media, shifting content‑moderation responsibilities to a broader community.
Agrawal holds a bachelor’s degree from IIT Bombay and a Ph.D. in computer science from Stanford; his appointment continues the recent trend of Indian‑origin executives leading major U.S. tech firms.
Dorsey cited three reasons for stepping down, emphasizing the need to avoid founder‑centric single points of failure and to give the company independent leadership.
Following the announcement, Twitter’s stock rose about 2 % on Nasdaq, and Dorsey will remain on the board until his term expires in May 2022.
Agrawal expressed optimism about building on the company’s achievements and delivering greater value to users and shareholders as Twitter aims to reach 315 million monetizable daily active users by the end of 2023.
21CTO
21CTO (21CTO.com) offers developers community, training, and services, making it your go‑to learning and service platform.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
