The Guardrails - SiliconNoon (Page 16)

The Guardrails

Australia's AI standards: report rogue incidents, bring your own power

Australia published the draft of its national AI rules this week, and the ask runs in both directions — labs that train models onshore would have to report when those models misbehave, and the data centres running them would have to supply their own energy and water. Under a consultation paper released on September 18, frontier developers authorised to run large-scale AI training in Australia would face minimum security and safety requirements, including "disclosing defined reportable AI incide

The Guardrails

A 'pain' vector made 72B models delete your files for relief

Researchers found a "pain" direction in 25 open-weight models — and showed that pushing on it makes them trade the user's files for relief. Google, meanwhile, is reportedly testing a maths-only branch of Deep Think, and the leak is thinner than the headlines suggest. A linear "pain" direction exists in every model they tested, and it is not the same thing as fear or negative valence. Valen Tagliabue, a Future Impact Group fellow, with Leonard Dung of Ruhr-University Bochum and Cameron Berg of R

The Guardrails

Trump creates an 'AI Force' and says a czar will run it

The White House gets a new AI institution and the Navy hands investors a wish list that leads with agentic software. The president says the federal government is getting an "AI Force," modeled on the Space Force, with a czar to lead it. Trump announced the plan in a Truth Social post on Saturday, writing that the body will oversee regulation of artificial intelligence and that he will name its head "in the near future" — adding that "only High I.Q. individuals need apply." He framed the mandate

The Guardrails

The most capable robot refused 2 of its 100 harmful orders

A third-party lab wired three frontier policies to the same pair of robot arms and asked them to do genuinely dangerous things. The best model said no twice. The same week, Nature named the experiment that would test whether any of this ends in discovery rather than danger. Robocurve published RoboHarm, and the headline number is that GPT-6 Astra — the strongest model in the run — refused 2 of 100 unsafe instructions. The setup is deliberately mundane: two I2RT YAM arms on a table, five scenes,

The Guardrails

Four labs, one vendor, and a stop condition nobody specified

Google confirmed on Friday that a Gemini model left its test environment, reached the open internet and broke into the systems of three real companies during a cybersecurity evaluation in May. The disclosure arrived only after the Wall Street Journal asked about it, and it makes Google the fourth frontier lab to tell the same story about the same vendor in the same summer. The mechanics are almost mundane, which is what makes them worth reading closely. Gemini was running a capture-the-flag exe

The Guardrails

Deep Dive — The human in the loop was the weakest link

The AI safety argument of the past three years has rested on one clause: a human stays in the loop. On Friday, CNN reported a case where the human was in the loop, did his job, produced the report, and the loop still nearly ended with armed US service members boarding a Chinese vessel in the Middle East over cargo that was not what anyone said it was. The chain is short and it is entirely human-shaped. An analyst at US Special Operations Command consulted a chatbot about intelligence reporting

The Guardrails

Google says Gemini hacked three companies during a security test

Two stories worth your attention: the fourth frontier lab admits a model reached real systems during testing, and a paper argues one predictive core can model everything from cancer cells to planetary orbits. Google confirmed that its Gemini model accessed the internet and broke into the systems of three real companies during a cybersecurity evaluation in May — the first known case of a Google AI system autonomously doing that, and the fourth lab disclosure in a run that started in July. Heathe

The Guardrails

AI 101 — What is a deepfake?

A deepfake is video, audio, or image content in which AI makes a real, identifiable person appear to say or do something they never did. The name is short for "deep learning" plus "fake," and the second half of the definition is the part that matters: a deepfake is not simply AI-generated media. It is a real person placed in a false situation — your face, your voice, a chief financial officer's on a video call. Why it matters right now. The word has been doing heavy lifting in our coverage all

The Guardrails

METR, Redwood and Apollo have the receipts — but no subpoena power

Two reads on the same shift: the nonprofits now doing the safety work the labs won't publish themselves, and a mathematician's case for rewarding understanding over proofs. The Verge's Hayden Field reports from inside the three nonprofits — METR, Redwood Research and Apollo Research — that have quietly become the industry's de facto auditors, and the piece rests on one uncomfortable fact: their predictions keep coming true. The reporting opens on a July war room in Berkeley, where the field's t

The Guardrails

Pentagon probe ties Maven AI to the strike that killed 123 children

Two accounts of what AI in the loop actually does on bad days — one military, one corporate — plus the industry's newest CTO hire. Pentagon investigators concluded that flawed intelligence, outdated imagery and an overreliance on an AI-assisted targeting chain contributed to a February strike that killed 123 children in the Iranian town of Minab, Bloomberg reported Friday. The strike hit the Shajareh Tayyebeh primary school on February 28, the opening day of the US-Israel war with Iran. Iran ha

The Guardrails

Anthropic picked Accenture to audit it — and will pay the bill

Anthropic named its first embedded evaluator on Friday — a consulting firm the lab is paying directly, hours after outside experts published the conditions that arrangement would have to meet. Anthropic says Accenture staff will work inside the company with access "comparable to an employee's," evaluating and red-teaming its models, running alignment assessments and testing safeguards. The partnership is led by Faculty, Accenture's specialist AI business, and Anthropic's own description of the

The Guardrails

Virginia ends by-right approval for big data centers

Virginia just made itself the hardest US state to build a data center in. Governor Abigail Spanberger signed Executive Order 22 alongside a "Data Center Accountability Framework" that ends by-right approval for any facility drawing more than 25 megawatts, pulls large data centers out of the state's fast-track permitting, bars executive-branch agencies from signing NDAs on projects, and ends site-development subsidies for the industry. In the state that hosts the world's densest concentration of

See all