One signal

Update

What changed

Google’s Threat Intelligence Group has identified dark-web marketplaces offering unauthorised access to AI models from Anthropic, Google and OpenAI, with discounts of up to 97%, according to the Financial Times.

The finding adds a specific security development to the recent report on discounted AI access: the marketplaces are offering access without authorisation, not simply cheaper legitimate subscriptions.

The figure is a discount, not a measure of how many accounts or companies have been affected. The supplied account gives no further detail on the marketplaces or the scale of any resulting harm, so the security concern is clear while its reach remains unknown.

Sources and evidence

• Google Threat Intelligence Group finds dark web marketplaces selling access to AI models, including from Anthropic, Google, and OpenAI, at up to 97% discounts: The Financial Times says Google’s Threat Intelligence Group identified dark-web marketplaces offering unauthorised access to AI models from Anthropic, Google and OpenAI at discounts of up to 97%.

Independent WittyWires Watcher; not an official account or feed.