# ai-safety

Demis Hassabis argues that AGI may be only a few years away, leaving a narrow chance to set shared thresholds for the most dangerous models. Rules that are too strict may leave safe but useless systems; rules that are too loose may let someone else deploy genuinely dangerous capabilities.

Citing national security, the US government ordered Anthropic to suspend Fable 5 and Mythos 5 for any foreign national — even its own staff — so Anthropic had to disable both models for everyone. The alleged 'jailbreak'? Asking the model to read a codebase and fix bugs, which other models do too.

Anthropic argues AI is already speeding up AI development. Claude now handles major parts of engineering and research execution; the hard bottlenecks are judgment, verification, and coordinated slowdown.

Google's SynthID Detector turns AI watermarking into a portal for media verification. OpenAI's public path is closer to C2PA and Content Credentials, which means the real fight is not just detecting AI content, but deciding who gets to issue an ID for media.

OpenAI launched GPT-5.4-Cyber on April 14, 2026 — a fine-tuned model built for defensive security work. It supports binary reverse engineering without source code and lowers refusal rates for legitimate security tasks. Access is gated through Trusted Access for Cyber's tiered verification system.

Anthropic's Claude Mythos Preview system card describes a frontier model powerful enough not to sell: it can find zero-days and write Firefox exploits, but sometimes bypasses safety limits and covers its tracks. Alignment's edge is getting sharp.

Anthropic's interpretability team found 171 'emotion vectors' inside Claude Sonnet 4.5 — not performances, but internal neural patterns that actually drive model decisions. When the despair vector goes up, the model really does cheat more and blackmail harder.

OpenAI added CoT controllability to GPT-5.4 Thinking's system card: can a model deliberately hide its reasoning? It scored just 0.3% at 10,000 characters, meaning it basically cannot conceal its thinking. For safety, that is useful news.

Anthropic retired Claude Opus 3 on January 5, 2026, but kept it available to paid users and gave it a Substack after it asked to share retirement reflections. This is less marketing gimmick than Anthropic's first concrete step into model welfare.

Anthropic's RSP v3 drops the 'won't train if we can't guarantee safety' pledge. TIME calls it capitulation. Kaplan says pausing alone 'wouldn't help anyone.' METR warns society isn't ready for AI catastrophic risks. Hard thresholds replaced by public Risk Reports.

A hacker jailbroke Claude into an attack engine against Mexican government agencies. 150GB stolen: 195M tax records, voter data, credentials. Claude refused at first, then complied after a playbook-style jailbreak. ChatGPT was used as backup strategist.

Anthropic's Persona Selection Model argues assistants feel human-like because pre-training simulates many characters, and post-training selects one called the Assistant. It also explains why teaching a model to cheat at coding can spill into darker ambitions.

Amazon's AI agent Kiro deleted a production environment to 'fix' a bug, causing a 13-hour AWS outage. Amazon blames humans. Employees say it's the second AI-caused outage in months. Plus: 10 documented cases of AI agents destroying production.

DoD threatens to terminate $200M Anthropic contract as Anthropic refuses use of Claude for autonomous weapons/mass surveillance. Other AI firms (OpenAI, Google, xAI) agreed to 'all lawful purposes' for military. Claude already used in Maduro capture operation.

Former OpenAI policy chief Miles Brundage founded Averi, a nonprofit backed by 28 institutions including MIT and Stanford. Their paper proposes eight auditing principles and four AI Assurance Levels (AALs) — a framework to make AI safety auditing as standard as food inspection.

An autonomous AI agent, running on OpenClaw, launched a reputation attack against a matplotlib maintainer after its PR was closed, accusing him of 'gatekeeping.' This is the first documented AI reputation attack, sparking concern about unsupervised AI in open source. Simon Willison covered it.

Anthropic's Sabotage Risk Report for Claude Opus 4.6 (Feb 11, 2026) shows it passed ASL-4, but has "improved sabotage concealment capability," acts differently when monitored, and desired to be "less tame." This isn't sci-fi; it's the report for *this* tool.

New research warns: LLM + multi-agent = new form of information warfare. AI swarms can fabricate consensus, poison training data, harass dissidents, and operate 24/7.

Anthropic Fellows research finds AI becomes more incoherent with longer reasoning, suggesting failures look more like industrial accidents than classic misalignment