OpenAI's Pentagon Deal: A Complex Web of Ethics and Politics
The recent deal between OpenAI and the US Department of Defense (DoD) has sparked intense debate and raised ethical concerns. The initial agreement, which was hastily arranged, faced scrutiny from the public and employees of both OpenAI and Google, who signed an open letter opposing the use of AI for surveillance and autonomous killing.
The deal's announcement came soon after OpenAI's ChatGPT, with over 900 million users, was chosen to supply AI to the Pentagon. This decision was made shortly after Anthropic, the previous AI contractor, was dropped due to concerns over its technology's potential for mass domestic surveillance. Anthropic's CEO, Sam Altman, had previously stated that using AI for such purposes was incompatible with democratic values.
However, the new deal faced immediate backlash. Critics pointed to the Snowden scandal, where the NSA was found to be harvesting phone and internet communications, raising fears that OpenAI's technology could be used for similar purposes. This led to an online campaign, with users urging a 'delete ChatGPT' movement, demanding transparency and accountability.
In response, OpenAI's CEO, Sam Altman, admitted that the deal's haste and lack of clear communication made it appear 'opportunistic and sloppy'. He emphasized the complexity of the issues and the need for careful consideration. The company's initial statement about the contract's 'guardrails' was questioned, as employees and observers raised concerns about the potential for misuse.
The open letter, signed by 796 Google employees and 98 OpenAI staff, highlighted the US government's attempt to divide companies with fear of each other's compliance. The letter called for unity in refusing the DoD's demands for domestic mass surveillance and autonomous killing without human oversight.
Miles Brundage, OpenAI's former head of policy research, expressed skepticism about the deal, suggesting that OpenAI may have caved to government pressure. He emphasized the importance of democratic processes and the need for government decisions to be guided by principles of liberty, rather than unconstitutional orders.
As a result, three US cabinet-level agencies have ceased using Anthropic's AI products, and Trump has ordered a phase-out of Anthropic's technology across all federal agencies, further complicating the AI industry's relationship with government contracts.
