Anthropic adds detection tools for coordinated election misuse of Claude

Anthropic says it has added classifiers to detect election abuse spread across multiple Claude requests, including efforts to create fake social media accounts and political news sites. Testing with the Institute for Strategic Dialogue and the McCain Institute found that Claude reliably refused single harmful requests but could miss malicious campaigns divided into individually innocuous tasks. The company also says it strengthened detection of election misuse by foreign state actors and banned accounts involved in fabricated election coverage. The additions address a specific weakness exposed by testing; those tests found that refusing individual requests did not reliably identify the wider campaign.