This week, five stories landed within hours of each other that, on the surface, look unrelated. Underneath, they\u2019re all asking the same question: now that AI can do the thing, who\u2019s in charge when it does?<\/p>\n
Let\u2019s start with Meta: Apparently, one of their AI agents went rogue, posted unauthorized analysis of company and user data on an internal forum, and triggered a Sev 1 security incident (the same severity level reserved for major outages and data breaches).<\/p>\n
Meta ran its employee-leak playbook on a piece of software. The agent wasn\u2019t hacked. It just\u2026 acted.<\/p>\n
Meanwhile, MiniMax shipped M2.7,<\/a> a model that ran 100+ optimization rounds on itself during training, scoring 56% on SWE-Bench Pro<\/a> (near-Opus level on coding) and at $0.30 per million tokens<\/a>. The model now handles 30-50% of MiniMax\u2019s research. The tool is building its next version. And the lab is letting it.<\/p>\n Then there\u2019s Xiaomi<\/a>. Their trillion-parameter model, MiMo-V2-Pro, sat on OpenRouter<\/a> for a week as \u201cHunter Alpha\u201d and the entire developer community attributed it to DeepSeek<\/a>. Nobody could identify the maker from the output. If you can\u2019t tell who built a frontier model by using it, competitive moats based on model quality are dissolving fast.<\/p>\n Apple\u2019s response to all of this? As mentioned above, block the vibe coding apps<\/em>. If anyone can prompt an app into existence, the App Store\u2019s developer ecosystem, the thing Apple\u2019s revenue model depends on, becomes optional.<\/p>\n Their answer: lock the gate.<\/p>\n And Anthropic<\/a> just surveyed 81,000 Claude users across 159 countries. The finding that should make everyone pause: hope and alarm aren\u2019t splitting people into opposing camps (pro vs anti AI). Actually, they coexist inside the same person. There\u2019s no clean \u201cpro-AI\u201d vs. \u201canti-AI\u201d debate to have. The tension is internal. Wow, it\u2019s almost like what we try to do here at The Neuron is the natural human response\u2026<\/em><\/p>\n Every one of these pieces is about what happens after<\/em> AI can do the thing.<\/p>\n Meta<\/a> can\u2019t control agents post-deployment.<\/p>\n MiniMax<\/a> can\u2019t fully predict what a self-improving model becomes.<\/p>\n Xiaomi proved you can\u2019t trace who built what.<\/p>\n Apple<\/a> is locking doors because users no longer need developers.<\/p>\n And 81,000 people are saying the real question isn\u2019t whether AI works; it\u2019s who decides what happens when it does.<\/p>\n And the scary part? Nobody has an answer yet. Everyone\u2019s improvising.<\/em><\/p>\n Gee, it\u2019s almost like we as a society should have a governing body of some sort that can pass rules we all agree to follow and then enforce them to help us decide what should and shouldn\u2019t happen next\u2026 too bad we APPARENTLY DON\u2019T?!<\/em><\/p>\n Wait a minute\u2026 signs of life! Signs of something happening: Apparently, Senator Blackburn<\/a> released a draft \u201cTrump America AI Act\u201d (full text<\/a>) that would codify the December 2025 AI executive order and preempt state AI laws<\/a> with a single national standard.<\/p>\n What else does this mean: The \u201ccapability\u201d question re: \u201ccan\u201d <\/em>a model do XYZ is settled. I mean, not really, but it basically is. As soon as an AI can even begin to do a thing, ppl will use it to do the thing. So whatever capabilities you don\u2019t have now or are concerned about happening later, just give it a few more rounds of model drops. We\u2019ll get there. Which is why it\u2019s so important we decide what they should \/ shouldn\u2019t be allowed to do.<\/em><\/p>\n Editor\u2019s note: This content originally ran in the newsletter of our sister publication, <\/strong>The Neuron<\/strong><\/a>. To read more from The Neuron, <\/strong>sign up for its newsletter here<\/strong><\/a>.<\/strong><\/p>\n The post This Week in AI: Meta Incident Highlights a Loss of Control<\/a> appeared first on eWEEK<\/a>.<\/p>","protected":false},"excerpt":{"rendered":" This week, five stories landed within hours of each other that, on the surface, look unrelated. Underneath, they\u2019re all asking the same question: now that AI can do the thing, who\u2019s in charge when it does? Let\u2019s start with Meta: Apparently, one of their AI agents went rogue, posted unauthorized analysis of company and user […]<\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[],"class_list":["post-7897","post","type-post","status-publish","format-standard","hentry","category-news"],"_links":{"self":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/7897"}],"collection":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=7897"}],"version-history":[{"count":0,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/7897\/revisions"}],"wp:attachment":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7897"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7897"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7897"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Why this matters<\/h2>\n