Sam Altman Leaves OpenAI's Safety Committee, Handing Over to Independent Group

According to the announcement on OpenAI's official blog, the Safety and Security Committee will become an independent oversight group under the board. Carnegie Mellon University professor Zico Kolter will chair the committee, alongside Quora CEO Adam D'Angelo, retired US Army General Paul Nakasone, and former Sony executive vice president Nicole Seligman. All of these members are currently members of the OpenAI board.

Following Altman’s resignation, the committee completed a safety review of the company’s latest AI model, the o1. The group will continue to receive briefings from OpenAI’s safety and security team, and reserves the right to delay the release if safety concerns are not addressed.

According to OpenAI, the Committee will maintain an important role in providing technical oversight for current and future models, along with post-release monitoring. The company is committed to building built-in safety processes with clear evaluation criteria to ensure that all of its AI models are released safely.

Altman’s decision to leave the committee comes after a group of five US senators sent a letter asking OpenAI to clarify its safety policies. Nearly half of OpenAI’s staff, which focused on studying the long-term risks of AI, has left the company in recent years. Former OpenAI researchers have criticized Altman for not advocating for strong AI regulation, instead favoring policies that advance the company’s commercial goals.

Image credit: Justin Sullivan / Getty Images

Additionally, OpenAI has significantly increased its spending on lobbying the US Congress . The company has budgeted $800,000 for the first six months of 2024, up from $260,000 for all of 2023. Altman also joined the US Department of Homeland Security's Artificial Intelligence Safety and Security Board, where he contributed to recommendations on AI across the country's critical infrastructure.

Although Altman has left the Security and Privacy Committee, there seems to be little indication that the committee will make decisions that could seriously impede OpenAI’s commercial path. In May, OpenAI said it would seek to address “valid criticism” of its work through the committee, although “valid criticism” is largely dependent on the perspective of the reviewer.

In an editorial in The Economist in May, two former OpenAI board members, Helen Toner and Tasha McCauley, expressed concerns about the company’s ability to govern itself, arguing that OpenAI is currently unable to hold itself accountable under the pressure of profit.

Profits are becoming an increasingly important driver for OpenAI. The company is in the process of raising $6.5 billion in a new round of funding that would value OpenAI at more than $150 billion. To attract investors, OpenAI may consider abandoning its hybrid nonprofit structure, which is designed to limit returns to investors in favor of its goal of developing AI for the common good of humanity.

Hung Nguyen (According to TechCrunch)

Source: https://www.congluan.vn/sam-altman-roi-uy-ban-an-toan-cua-openai-nhuong-quyen-cho-nhom-doc-lap-post312681.html