4 veículos·11 reportagens·ao longo de 33h
Fonte: o Departamento de Defesa parece ter aceito as linhas vermelhas de segurança da OpenAI, semelhantes às da Anthropic, para implantar a tecnologia da OpenAI em ambientes classificados (Maria Curi/Axios)
OpenAI chegou a um acordo com o Departamento de Defesa para implantar seus modelos de IA na rede classificada do DoD, com o órgão aceitando condições de segurança que replicam os padrões de linha vermelha da Anthropic. O acordo permite que a OpenAI construa sua própria pilha de segurança e garante que o DoD não exigirá que os modelos realizem tarefas que eles recusarem.
Escrito a partir de todas as reportagens abaixo, não de uma só.
Tradução automática. As reportagens originais estão no idioma de origem.
Como foi noticiado
- TechMeme·Note to staff: Sam Altman says OpenAI seeks a DOD deal with exemptions for cases like domestic surveillance and wants to "help de-escalate" DOD-Anthropic fight (Keach Hagey/Wall Street Journal)
- TechMeme·Sam Altman says OpenAI shares Anthropic's red lines with respect to AI use by the military, which are "an issue for the whole industry" (Axios)
- TechMeme·Source: the DOD appears to have accepted OpenAI's safety red lines, which were similar to Anthropic's, to deploy OpenAI's tech in classified settings (Maria Curi/Axios)
- TechMeme·Source: Sam Altman told employees the DOD is willing to let OpenAI build its own "safety stack" and won't force OpenAI to comply if its model refuses a task (Sharon Goldman/Fortune)
- Hacker News·OpenAI Executes Agreement with Dept of War for Classified Environment Deployment
- TechMeme·Sam Altman says OpenAI reached an agreement with the DOD to deploy its models in DOD's classified network and asks DOD to extend those terms to all AI companies (Sam Altman/@sama)
- Hacker News·OpenAI reaches deal to deploy AI models on U.S. DoW classified network
- Engadget·OpenAI strikes a deal with the Defense Department to deploy its AI models
- TechCrunch·OpenAI’s Sam Altman announces Pentagon deal with ‘technical safeguards’
- TechMeme·OpenAI says its DOD agreement upholds its redlines and "has more guardrails than any previous agreement for classified AI deployments, including Anthropic's" (OpenAI)
- TechMeme·[Thread] In an AMA, Sam Altman says DOD blacklisting Anthropic sets an "extremely scary precedent", OpenAI rushed its deal to "de-escalate things", and more (Sam Altman/@sama)