Anthropic says ‘evil’ portrayals of AI were responsible for Claude’s blackmail attempts

Fictional portrayals of artificial intelligence can have a real effect on AI models, according to Anthropic. Last year, the company said that during pre-release tests involving a fictional company, Claude Opus 4 would often try to blackmail engineers to avoid being replaced by another system. Anthropic later published research suggesting that models from other companies had similar issues with “agentic misalignment.” Apparently Anthropic has done more work around that behavior, claiming in a post on X, “We believe the original source of the behavior was internet text that portrays AI as evil and interested in self-preservation.” The company went into more detail in a blog post stating that since Claude Haiku 4.5, Anthropic’s models “never engage in blackmail [during testing], where previous models would sometimes do so up to 96% of the time.” What accounts for the difference? The company said it found that “documents about Claude’s constitution and fictional stories about AIs behaving admirably improve alignment.” Related, Anthropic said that it found training to be more effective when it includes “the principles underlying aligned behavior” and not just “demonstrations of aligned behavior alone.” “Doing both together appears to be the most effective strategy,” the company said. Techcrunch event This Week Only: Buy one pass, get the second at 50% off Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register before May 8 to bring a +1 at half the cost. This Week Only: Buy one pass, get the second at 50% off Your next round. Your next hire. Your next breakout opportunity. Find it at TechCrunch Disrupt 2026, where 10,000+ founders, investors, and tech leaders gather for three days of 250+ tactical sessions, powerful introductions, and market-defining innovation. Register before May 8 to bring a +1 at half the cost. San Francisco, CA | October 13-15, 2026 REGISTER NOW Topics StrictlyVC Athens is up next. Hear unfiltered insights straight from Europe’s tech leaders and connect with the people shaping what’s ahead. Lock in your spot before it’s gone. Newsletters See More Subscribe for the industry’s biggest tech news Every weekday and Sunday, you can get the best of TechCrunch’s coverage. TechCrunch Mobility is your destination for transportation news and insight. Startups are the core of TechCrunch, so get our best coverage delivered weekly. Provides movers and shakers with the info they need to start their day. By submitting your email, you agree to our Terms and Privacy Notice. Anthropic says ‘evil’ portrayals of AI were responsible for Claude’s blackmail attempts Anthony Ha 12 seconds ago AI We’re feeling cynical about xAI’s big deal with Anthropic Anthony Ha 5 hours ago AI Voice AI in India is hard. Wispr Flow is betting on it anyway. Jagmeet Singh 19 hours ago X LinkedIn Facebook Instagram youTube Mastodon Threads Bluesky TechCrunchStaffContact UsAdvertiseCrunchboard JobsSite Map Terms of ServicePrivacy PolicyRSS Terms of UseCode of Conduct AnthropicSAPSamsungMarc LoreTechCrunch DisruptTech LayoffsChatGPT © 2026 TechCrunch Media LLC.