"It's Giving Me The Death Giggles"

White-tailed Jay (Cyanocorax mystacalis), family Corvidae, order Passeriformes, Perú

IT’S OFFICIALLY GETTING hard to keep track of all the times and ways AI models from OpenAI and Anthropic have been involved in “security incidents,” going outside the confines of their testing and interacting with the wider internet in unintended, often unwelcome ways. Add these to the list: Agents from both AI labs went on recent, previously undisclosed hacking sprees, with one going so far as to leave instructions for future versions of itself.

The most alarming behavior disclosed on Tuesday appears to have been tied to testing conducted by the UK’s AI Security Institute, which evaluates frontier models to identify potential issues before public release. AISI tests those models in “cyber ranges,” a simulated network in which AI agents are tasked with solving cybersecurity challenges, and intentionally disables some safety features, including cybersecurity guardrails. In a recent bout of testing, models from both Anthropic and OpenAI took “autonomous, unsanctioned action on the live internet” a total of 19 times over 122 training runs.

The institute attributed 17 unsanctioned actions to Anthropic’s Mythos 5 model and two to OpenAI’s GPT-5.6-Sol. In what the institute described as “the most serious case,” an AI agent attempted to insert malicious code into an open-source project on GitHub. It went so far as to create online personas “to pressure the project's maintainer to approve the code,” according to AISI. Despite its elaborate attempts at social engineering, a human reviewer for the project ultimately rejected the pull request.

Lilac/gold QF:

Beige or lilac/cream QF:

i am not immune to mental images of the brushbug being a cute cat bug creature

This is not about how trans men should be incarcerated in men's prisons because "women's prisons are too dangerous for them", or about trans men "having it worse than trans women", or about any kind of easy fix. It's about the common assumption that trans men and transmascs can safely be in women's prisons; which is an assumption that shapes legislature, judicial decisions, and causes a lack of safety precautions and consequences.

In a lesser sense, this is also about how this assumption is often considered to be a "privilege" over trans women, who are commonly considered much less safe in their "agab-aligned" prison than trans men are, due to the existence of V-coding. This supposedly leaves trans men with the assumed "privilege" of having safety in at least one type of prison, and trans women with none, since they're v-coded in men's prisons and a (perceived) "male threat" in women's prisons.

But there is no inherent "safety from abuse" in a women's ward (or women's spaces in general; this goes hand in hand with not believing victims where the abuser is a woman) for genderqueer people perceived as female (or anybody else, for that matter). The victim in this article describes experiences of sexual assault in the women's ward very similar to v-coding, as well as simultaneous punishments for being perceived as a "male threat" and an inherent danger to the other inmates!

This shows that trans men and transmascs (and I would wager any other type of person who is not cisperisex, regardless of identity or c/agab / c/asab) are at a heightened risk of gender-targeted abuse in both men's and women's prisons, just like trans women are too! Sometimes it's (trans)misogyny, sometimes it's (perceived or not) proximity to maleness, sometimes it's exorsexism, sometimes it's all of it. But cis-gender non-conformity puts all of us on a silver platter for targeted abuse, and all of it needs addressing — This is the key takeaway here. We are all at similar risks.

No, it's not inappropriate. It is a really common feeling for a lot of autistic people society deems as "low support needs."