OpenAI announced today that it is releasing Sol, its most capable AI model to date, to a select group of about 20 partners. Each partner’s participation was vetted and approved by the U.S. government, a step that follows a direct request from the Trump administration to stagger the model’s debut.

Sol tops a new three‑tier lineup that also comprises Terra, a mid‑range offering, and Luna, which emphasizes speed and cost efficiency. OpenAI describes Sol as excelling in coding, biology and cybersecurity tasks. The model introduces a “max reasoning effort” mode that allows it to spend extra time on complex problems, and the company says an “ultra” mode that distributes work across multiple sub‑agents is in development.

The limited preview is the first practical test of the AI executive order signed in early June. That order asks AI firms to voluntarily grant the government up to 30 days of pre‑release access to models with advanced cyber capabilities, while explicitly rejecting mandatory licensing. By approving each partner individually, the government is applying a managed‑access list that goes beyond the order’s original voluntary review framework.

OpenAI’s blog post stresses that the company does not view this government‑gated process as a long‑term default, but it agreed to cooperate to avoid a more coercive outcome. The decision contrasts with recent actions against rival Anthropic, which was ordered to shut down its Fable 5 and Mythos 5 models after a reported jailbreak. Anthropic complied but labeled the move disproportionate and warned it could halt future frontier model deployments if applied broadly.

Sol is also the first model in the series to be offered through Amazon Bedrock, marking a rare instance of a cutting‑edge OpenAI model appearing on a competing cloud platform. OpenAI says it intends to make all three tiers generally available in the coming weeks, though no specific public launch date has been set.

The rollout raises questions about whether government‑managed releases will become the template for future frontier AI models. OpenAI appears intent on keeping the current arrangement temporary, positioning voluntary compliance as a way to preserve flexibility while meeting regulatory expectations.

Este artículo fue escrito con la asistencia de IA.
News Factory APP - noticias agénticas para impulsar tu SEO y AEO.