Anthropic

If you decide to turn off the model training setting, we will not use any new chats [...] for future model training. Unverified source (2025)

Importantly, the law balances [...] whistleblower protections—[...]. [...] Protecting whistleblowers: It should be an explicit violation of law [...] punish employees who raise concerns [...]. Unverified source (2025)

we propose that AI developers pre-register large training runs [...] within their home country’s government [...] plausible bar for now would be 10^26 FLOP or higher. Unverified source (2023)

The definition, criteria, and safety measures for each ASL level are described in detail in the main document, but at a high level, ASL-2 measures represent our current safety and security standards and overlap significantly with our recent White Hou... more Unverified source (2023)

Release public transparency reports summarizing their catastrophic risk assessments and the steps taken to fulfill their respective frameworks before deploying powerful new models. Unverified source (2025)

for example ability to give instructions on how to build bioweapons Unverified source (2023)