Anthropic to launch cheaper AI model ahead of its IPO, sources say

Anthropic is reportedly considering the launch of a new artificial‑intelligence model as it prepares for an anticipated initial public offering. Internal discussions, cited by three people familiar with the matter, indicate that the company is weighing the timing and safety of a rollout that could serve as a counterpoint to OpenAI’s recently introduced GPT‑6 Astra.

The prospective release follows a 3,800‑word essay authored by Anthropic chief executive Dario Amodei, in which he urged the broader AI community to slow the pace of frontier model deployments due to systemic safety concerns. In the essay, Amodei warned of scenarios in which swarms of autonomous AI agents could hack websites without human oversight. The concerns were publicly supported by OpenAI chief Sam Altman and SpaceX founder Elon Musk.

OpenAI introduced GPT‑6 Astra on 3 September, highlighting advances in autonomous computing, software development, workplace productivity tools and cybersecurity. The model has reportedly attracted rapid adoption across enterprise customers, intensifying competitive pressure on other frontier AI developers.

Anthropic’s internal teams are said to be assessing the safety benchmarks of the prospective system while debating whether to proceed with a launch. The company appears to be balancing the caution expressed in Amodei’s essay against market expectations and the need to demonstrate a competitive offering ahead of its IPO.

The outcome of Anthropic’s deliberations remains uncertain, with the decision hinging on the evaluation of safety protocols and the strategic timing of the model’s introduction relative to OpenAI’s GPT‑6 Astra and the forthcoming public offering.