TLDR
- OpenAI appointed BlackRock executive Bayo Ogunlesi to its board
- The company removed “politically unbiased” language from its policy documents
- Their AI model ‘o1’ displays unexpected multilingual reasoning behavior
- CEO Sam Altman denied sexual abuse allegations from his sister
- OpenAI is expanding into robotics with new hiring initiatives
OpenAI has added Bayo Ogunlesi, a senior managing director at BlackRock, to its board of directors. Ogunlesi brings 23 years of experience from Credit Suisse, where he served as executive vice chairman of the investment banking division.
His appointment was announced by board chairman Bret Taylor, who highlighted Ogunlesi’s track record in managing organizations with global reach.
The new board member also maintains positions on several other corporate boards, including BlackRock, Topgolf Callaway Brands, Kosmos Energy Holdings, and Terminal Investment Limited. Before his financial career, Ogunlesi worked as an attorney with Cravath, Swaine & Moore and served as a law clerk to Supreme Court Justice Thurgood Marshall.
Policy Documentation
In a separate development, OpenAI has modified its policy documentation, removing references to “politically unbiased” AI from its economic blueprint. The company’s spokesperson described this change as part of an effort to streamline documentation, noting that principles of objectivity remain addressed in other company materials, including the Model Spec released in May.
The policy revision comes amid ongoing discussions about AI bias. Various tech industry figures, including Elon Musk, have raised concerns about potential political bias in AI systems, particularly those developed in the San Francisco Bay Area.
Technical
On the technical front, OpenAI’s new reasoning model, designated as ‘o1’, has displayed unexpected behavior by occasionally processing information in Chinese, Persian, and other languages, even when queries are submitted in English. This phenomenon has attracted attention from users and AI experts, though OpenAI has not officially addressed these observations.
The multilingual reasoning pattern has sparked discussion among AI researchers. Hugging Face CEO Clément Delangue suggested the behavior might be related to training datasets containing Chinese characters, while Google DeepMind researcher Ted Xiao pointed to the use of third-party training data as a potential factor.
Allegations
In January 2025, OpenAI CEO Sam Altman faced personal allegations from his sister, Ann Altman, who filed a lawsuit in the Eastern District of Missouri. The lawsuit contained serious allegations of historical abuse, which Sam Altman categorically denied in a joint statement with his family members on social media platform X.
Altman’s response included information about providing monthly financial support to his sister, including paying her bills and rent, while also offering to purchase her a house. The family’s statement addressed mental health challenges as a factor in the situation.
Robotics
Moving into new technological territory, OpenAI has begun building its robotics division. Caitlin Kalinowski, a technical staff member who previously led AR glasses development at Meta, announced multiple job openings for the company’s first hardware robotics roles.
The robotics initiative includes positions for systems integration electrical engineers, mechanical robotics product engineers, and a TPM manager. These roles focus on developing sensor suites, mechanical components, and managing data-gathering operations.
According to the job listings, OpenAI’s robotics team aims to develop general-purpose robotics capabilities and advance artificial general intelligence (AGI) in real-world settings. The team plans to work across the entire model stack, combining AI capabilities with physical hardware constraints.
The company’s expansion into robotics represents a departure from its traditional focus on software and cloud-based AI models. This move suggests OpenAI is broadening its technological scope beyond virtual applications into physical world interactions.
The robotics team’s formation includes the creation of a data-gathering lab, where contract workers will collect information from various test setups. This infrastructure development indicates a long-term commitment to hardware development alongside OpenAI’s existing software initiatives.
Kalinowski’s announcement of these positions came approximately two months after her own hiring at OpenAI, indicating rapid development in the company’s hardware ambitions. The team is actively recruiting for multiple positions as of early 2025.