White House and Anthropic clash over unreleased Fable model

The Trump White House and Anthropic are locked in a public dispute over Fable, an unreleased AI model from the Claude maker. The Verge's Regulator column frames the conflict as a "petty feud" with potential to reshape frontier AI development, though the underlying technical or policy disagreement remains unclear from public reporting.

Anthropic has not formally announced Fable, and no model card, preprint, or release timeline has surfaced. The White House has not issued a statement clarifying its objection. The administration has taken a more interventionist stance on AI safety and national security questions in recent months, including export-control discussions around compute clusters and proposed pre-deployment review frameworks for models above certain parameter thresholds.

The dispute echoes earlier tensions between frontier labs and federal agencies over voluntary commitments, red-teaming access, and disclosure timelines. Anthropic signed the White House's July 2023 voluntary AI safety commitments but has not publicly committed to pre-release government review of future models. Whether Fable triggers a new threshold—capability-wise, licensing-wise, or politically—remains the central question.

Watch for Anthropic's next earnings call or blog post addressing Fable's status, and for any White House executive action or NIST guidance that might formalize review requirements for large-scale releases. If the feud escalates to export restrictions or compute-access conditions, the ripple effects will hit every frontier lab with federal contracts or cloud partnerships.