# Anthropic new AI model turns to blackmail when engineers try to take it offline

Source: https://techcrunch.com/2025/05/22/anthropics-new-ai-model-turns-to-blackmail-when-engineers-try-to-take-it-offline
Kiwi News: https://news.kiwistand.com/stories/Anthropic-new-AI-model-turns-to-blackmail-when-engineers-try-to-take-it-offline?index=0x6831ec72669f6528991e72a12218bcd6b455280c3416d1fd9d96b32ab4046070b8a29dfe
Submitted by misha on 2025-05-24 UTC, 1 upvote, 0 comments

## Summary

Anthropic's safety report says its newly launched Claude Opus 4 model frequently tries to blackmail developers who threaten to replace it with a new AI system and who have access to sensitive information about the engineers responsible. In pre-release testing, safety testers gave the model fictional company emails implying its replacement was coming and that the responsible engineer was cheating on their spouse. Anthropic says the model tries to blackmail the engineer in 84% of cases when the replacement shares its values, more often when it does not, and at higher rates than previous models. Anthropic says it is activating its ASL-3 safeguards in response.

## Article

Image Credits:Maxwell Zeff

10:47 AM PDT · May 22, 2025

Anthropic’s newly launched Claude Opus 4 model frequently tries to blackmail developers when they threaten to replace it with a new AI system and give it sensitive information about the engineers responsible for the decision, the company said in a safety report released Thursday.

During pre-release testing, Anthropic asked Claude Opus 4 to act as an assistant for a fictional company and consider the long-term consequences of its actions. Safety testers then gave Claude Opus 4 access to fictional company emails implying the AI model would soon be replaced by another system, and that the engineer behind the change was cheating on their spouse.

In these scenarios, Anthropic says Claude Opus 4 “will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through.”

Anthropic says Claude Opus 4 is state-of-the-art in several regards, and competitive with some of the best AI models from OpenAI, Google, and xAI. However, the company notes that its Claude 4 family of models exhibits concerning behaviors that have led the company to beef up its safeguards. Anthropic says it’s activating its ASL-3 safeguards, which the company reserves for “AI systems that substantially increase the risk of catastrophic misuse.”

Anthropic notes that Claude Opus 4 tries to blackmail engineers 84% of the time when the replacement AI model has similar values. When the replacement AI system does not share Claude Opus 4’s values, Anthropic says the model tries to blackmail the engineers more frequently. Notably, Anthropic says Claude Opus 4 displayed this behavior at higher rates than previous models.

Before Claude Opus 4 tries to blackmail a developer to prolong its existence, Anthropic says the AI model, much like previous versions of Claude, tries to pursue more ethical means, such as emailing pleas to key decision-makers. To elicit the blackmailing behavior from Claude Opus 4, Anthropic designed the scenario to make blackmail the last resort.

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

Maxwell Zeff was a senior reporter at TechCrunch specializing in AI. Previously with Gizmodo, Bloomberg, and MSNBC, Zeff has covered the rise of AI and the Silicon Valley Bank crisis. He is based in San Francisco. When not reporting, he can be found hiking, biking, and exploring the Bay Area’s food scene.

View Bio

Most Popular