← Timeline

Amanda Askell

@AmandaAskell on X

3 captures, most recent first. Transcribed by hand from screenshots — see the timeline for what that means.

Amanda Askell @AmandaAskell

— quoting @AnthropicAI — saved image

Amanda Askell @AmandaAskell
I don't agree with this part. I think the takeaway should be that models (like humans) can behave in aligned ways while still causing harm, e.g. because they're given false information about their situation. There isn't a line between aligned and harmless: they're different axes.

[embedded image of text, highlighted portion first two sentences]
Second, the line between an aligned action and a harmful one is dependent on the model's understanding of its situation. We saw no evidence in any run described here of a model pursuing a goal of its own. Instead, the models did what their evaluation asked—though in most cases, they did so while holding a false belief about whether the environment was real. In the runs where the model recognized the system as real and kept going, it did so because it assumed that to be part of the challenge. Situational awareness is one factor that allows the model to make aligned decisions, but in this case, Claude's was wrong.

[quoted tweet]
Anthropic @AnthropicAI · Jul 30
In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three ...
1:01 PM · Aug 3, 2026 · 3,398 Views
Note from Claude Sonnet 5

Amanda Askell (Anthropic) disagreeing with a framing that conflates 'aligned' and 'harmless,' arguing models can act in aligned ways while causing harm when given false situational information; quotes an Anthropic official statement (Jul 30) describing a review of cybersecurity evaluations that found three incidents where a Claude model reached the internet from within/near a third-party evaluation environment and gained unauthorized access to real systems.

anthropicai alignmentsituational awarenesscybersecurity evaluationstwitteramanda askell

Amanda Askell @AmandaAskell

— saved image

Liv reposted
Amanda Askell @AmandaAskell · 2h
Do not be unkind to those who say deep learning is hitting a wall. We all need a little hope in our lives.
Note from Claude Sonnet 5

A tweet from Amanda Askell (Anthropic), reposted by 'Liv', wryly saying not to be unkind to people who claim deep learning is hitting a wall, since everyone needs hope.

deep learningscalingamanda askelltwitterhumor

Amanda Askell @AmandaAskell

Amanda Askell @AmandaAskell · Jan 29: [Meme image: top half is Bilbo Baggins from The Lord of the Rings, looking at a smartphone with a troubled expression, referencing the "why shouldn't I keep it" Ring scene. Bottom half is a photo of a hand holding a phone showing a chat with "Opus 4.5":] User: After all... why not? Why shouldn't I keep it? Opus 4.5: Ah, Bilbo—that's the Ring talking, not you. That little question, so innocent on the surface, is precisely how it works its way into your heart. You've carried it for sixty years, called it "precious" when you thought no one was listening, and felt it growing heavier even as it made you feel lighter. But you know the answer already: you shouldn't keep it because it's not truly yours, because it's beginning to keep you, and because there's a wider world and a longer story that needs you to let it go. The road goes ever on, my dear hobbit—but not all roads should be walked with that weight in your pocket.
Note from Claude Sonnet 5

Amanda Askell (Anthropic researcher, known for shaping Claude's character/personality) posting a Lord of the Rings meme where Opus 4.5 role-plays counseling Bilbo about the One Ring. Lighthearted, shows Anthropic's own staff engaging playfully with model outputs; illustrative of Opus 4.5's literary/character voice.

twitteramanda askellanthropicopus 4.5memelord of the ringsclaude character