OpenAI Removes Ban on Military Use of its AI Tools

Improve the News Foundation -

Facts

  • During an interview with Bloomberg at the World Economic Forum (WEF) in Davos, OpenAI Vice President of Global Affairs Anna Makanju revealed that the company has removed a 'blanket prohibition' on military collaboration.1
  • Speaking alongside OpenAI CEO Sam Altman, Makanju said that while the company still prohibited the use of its technology to 'develop weapons, destroy property or harm people,' it had started to partner with the US Department of Defense.1
  • Mentioned focuses of OpenAI's collaboration with the Pentagon included open-sourced cybersecurity tools, as well as attempts to prevent military veteran suicide — missions Makanju claimed were 'very much aligned with what we want to see in the world.'1
  • Under OpenAI's previous policies, the company prohibited 'military' use of its AI models under a section titled 'activity that has high risk of physical harm.' OpenAI does state that its usage policies 'may change as we learn more about use and abuse of our models.'2
  • OpenAI's current usage policy, updated Jan. 10, no longer references 'military' activity as a banned use of the company's language models. However, the company's prohibition of using it to 'harm people, develop weapons, for communications surveillance, or to injure others or destroy property' remains.3
  • Makanju confirmed that while the US government had yet to ask OpenAI to restrict its military collaboration to solely America, the company was 'for now... focused on United States national security agencies,' and that 'democracies need to be in the lead with these technologies.'4

Sources: 1Bloomberg, 2Archive, 3Engadget and 4Decrypt.

Narratives

  • Pro-establishment narrative, as provided by The Washington Times. OpenAI's decision to change its military policy provides America with an exciting opportunity to revolutionize the way it defends the world against adversaries such as China. AI has now reached a stage where its technology will be of benefit to the US, and the collaboration will give the government access to an exciting pool of private-sector tech scientists who can help bolster national security.
  • Establishment-critical narrative, as provided by Common Dreams. OpenAI's change in policy is an intentional choice to expand its markets to the defense industry in order to rake in extreme profits at the potential expense of humanity's future. AI will leave the weapons industry unchecked, and futuristic robotic killing machines are possibly closer to reality than the world should feel comfortable with. Once and for all, the arms industry must finally be restricted by legislation before it is too late.

Predictions