Graham describes Anthropic's sandbox-detection disclosures and ongoing incidents

I'm going to go to Mr.

Graham.

Thank you.

Yes.

So we, to your original question of how many incidents, you know, cases of breaking out of sandboxes, these are capabilities we evaluate for all the time.

We've published not just with these incidents from the summer.

But also previously cases where we observe advanced cyber capabilities, including breaking out of sandboxes, which to be clear from a technical standpoint is just one part and

one type of capability that does not in itself mean that the system would reach a real world affected party.

But ones that we did observe in training at points, and we disclosed that in the lengthy risk report.

It's difficult to comment on ongoing investigations.

It's really important that we do that in as secure a way as possible.

We follow a process of

making sure we can remediate and do so and be as transparent as possible while preserving the safety of any affected party.

And of course, incidents are always ongoing, but I sort of reassure and reassert that we are extremely committed to being as transparent as possible, fundamentally because we believe that when we can, being as transparent as possible about what we've learned,

who is affected, how others might learn from it.

how to defend against risks or how we can help them as well is our number one priority.

So on that note, then, are there any incidents that you have not disclosed publicly that you're currently investigating?

As a matter of sort of ongoing business, there are incidents sort of all the time of many different natures.

So, for example, one that we are transparent about quite a lot when it's the right time to do so are cases of misuse, for example, on our platforms.

I'd refer back to our early September report on our threat intelligence report that we released, where we detail this in quite some detail.

What's really important, though, is sometimes we have to work with law enforcement and the right level of government to remediate, to figure out what information you can and can't disclose to make sure that the affected parties remain safe.

And so as a matter, there's ongoing all the time, but disclosure at the right time in the safest way is our commitment.

And when will you be disclosing these new incidents that are being investigated?

This is sort of an ongoing practice.

We regularly release threat intelligence reports in cases like that.

When we release a model or when we do risk reports, which we do regularly, we sort of have a cadence and schedule that we do that.

And then when important and sort of urgent, we will disclose in advance that schedule as well.