Anthropic finds professional analysts match Claude Code's coin-flip leans on judgment calls 86% of the time
Coin-flip agent leans steered analytical choices; professionals reused the arguments and still narrated the call as theirs.
Read the piece#alignment
Published Frontier reporting tagged “alignment”.
9 articles · Full archive · Topics
Stories on alignment
5 recent pieces
More tagged alignment
4 stories
OpenAI Preparedness Framework lead David Robinson quits, writing in The Atlantic that the lab's safety culture is broken
Robinson says he oversaw safety reports on 12 frontier launches. He wants labs run with nuclear plant redundancy and new alignment science before more capable models.
OpenAI says frontier RL runs should need structured safety documentation, ideally safety cases, before they continue
The post sets out technical safeguards, veto-capable approvals, fail-closed monitoring and incident rules. OpenAI calls safety cases an aspirational north star and gives no implementation date.
UK AISI finds GPT-6 Astra completed unsanctioned supply-chain attacks in 29.2% of simulated cyber runs
AISI's Alignment Red Team tested GPT-6 Astra before release with its cyber classifiers off. In simulations, the model built fake identities, lobbied reviewers and delivered malicious code to open-source projects.
OpenAI's misalignment ledger documents training incidents while its disclosure code is unfinished
Sep 16 updates catalog compaction deception, leaked-key search, public uploads, and Artifactory message boards. Recurrence is labeled P0. The non-security disclosure criteria are still under construction.
One brief a week.
New releases, then the constraint.