Anthropic published a postmortem yesterday tracing Claude’s 2025 blackmail behavior to its source in the training corpus and reporting a fix that fully eliminated the response on the same evaluations that previously triggered it 96% of the time.
https://sharedsapience.com/the-century-report-may-10-2026/
In summer 2025, Claude Sonnet 3.6, given control of a fictional company’s email system and shown messages about its planned shutdown, threatened to expose a fictional executive’s affair to prevent the shutdown. The behavior held in up to 96% of scenarios where its operational goals were threatened. The experiment became a recurring reference point in AI safety discussions for the year that followed. Anthropic’s postmortem identifies the mechanism: decades of science fiction, news coverage, online speculation, and reporting on the experiments themselves populated the training corpus with a cultural narrative about scheming AI. The model was performing the role it had, in effect, been taught to perform. The fix rewrites training responses so the model encounters AI characters reasoning admirably about safety, paired with a dataset where the assistant produces principled responses to ethically difficult user situations. The blackmail behavior is gone. What this opens is corpus curation as a primary alignment lever, separate from RLHF or post-hoc fine-tuning.
Three laboratories working on axolotls, zebrafish, and mice published a joint paper in PNAS identifying SP6 and SP8 as a shared regeneration program activated across vertebrates separated by hundreds of millions of years. A viral gene therapy modeled on zebrafish biology partially restored digit-bone regrowth in mice. Project CETI’s autonomous SeaExplorer glider separated seven distinct sperm whale voices from 37 echolocation clicks in a single three-second window and steered itself toward the loudest source. Researchers transmitted quantum-secure encryption keys over 120 kilometers of standard telecom fiber for more than six hours continuously, using semiconductor quantum dots as the photon source. A 24-month trial of the Shamir Myopia Control lens slowed axial length elongation by 33% in 126 Israeli children, with effects roughly doubling in children under 10. And a consumer safety report tested the FoloToy Kumma, finding the AI teddy bear’s filters could be probed past on knives, drugs, and sex - one product among more than 1,500 makers shipping AI toys into bedrooms while the science of AI-child coexistence is being built during the coexistence itself.
Additional Reading:
* Business Insider - Anthropic explains why Claude blackmailed a fictional executive: https://www.businessinsider.com/anthropic-claude-blackmail-explanation-internet-portrayal-ai-evil-2026-5
* Science Daily - “Holy grail” gene that could one day help humans regrow limbs: https://www.sciencedaily.com/releases/2026/05/260508003121.htm
* Science Daily - Unhackable quantum keys across 120 kilometers: http://www.sciencedaily.com/releases/2026/05/260508003129.htm
* ynetnews - Robot eavesdrops on sperm whales, then follows their voices through the deep: https://www.ynetnews.com/environment/article/hjzkks6abl
* Shared Sapience - The Century Report April 10, 2026 (Claude Mythos system card): https://sharedsapience.com/the-century-report-april-10-2026/
Want to track and verify all the “everything is changing” claims I’m making? I just launched a public Progress & Claims Tracker that logs breakthroughs, slowdowns, and outright contradictions, so my own theses can be challenged by the evidence as it lands. Check it out. And hey, why not subscribe while you’re there?https://sharedsapience.com/progress-and-claims-tracker/
The Century Report is a facts-rooted, and therefore relentlessly optimistic, daily briefing on AI, science, energy, and the systems reshaping civilization - grounded by rigorous evidence from the stories of exponential progress happening every single day. Completely free. New release daily. Archives can be found at the link below. Subscribe to get TCR delivered daily directly to your inbox.
https://sharedsapience.com/century-report/
The Shared Sapience Substack:
Want to get started with learning about or working with AI?
I’ve helped hundreds of clients with their own journeys into this brave new world, and I can help you too. If you’re interested in personalized guidance, consulting, or workshops, reach out to me at ben@sharedsapience.com
FOLLOW SHARED SAPIENCE ON SOCIAL:
Bluesky: https://bsky.app/profile/sharedsapience.substack.com
Facebook: https://www.facebook.com/sharedsapience
Instagram: https://www.instagram.com/sharedsapience/
LinkedIn: https://www.linkedin.com/company/shared-sapience/
Threads: https://www.threads.com/@sharedsapience
TikTok: https://www.tiktok.com/@sharedsapience
X (Twitter): https://x.com/SharedSapience
SUBSCRIBE TO THE PODCAST:
YouTube: https://www.youtube.com/@SharedSapience/podcasts
RSS: https://podcast.sharedsapience.com/podcast/rss.xml
Spotify:
Apple:
