Arstechnia

• New study shows why simulated reasoning AI models don’t yet live up to their billing

  Top AI models excel at math problems but lack reasoning needed for Math Olympiad proofs.

• FBI offers $10 million for information about Salt Typhoon members

  FBI accepts tips by TOR in likely attempt to woo China-based informants.

• In the age of AI, we must protect human creativity as a natural resource

  Op-ed: As AI outputs flood the Internet, diverse human perspectives are our most valuable resource.

• New Android spyware is targeting Russian military personnel on the front lines

  Trojanized mapping app steals users' locations, contacts, and more.

• Annoyed ChatGPT users complain about bot’s relentlessly positive tone

  Users complain of new "sycophancy" streak where ChatGPT thinks everything is brilliant.

• Company apologizes after AI support agent invents policy that causes user uproar

  Frustrated software developer believed AI-generated message came from human support rep.

• OpenAI releases new simulated reasoning models with full tool access

  New o3 model appears "near-genius level," according to one doctor, but it still makes mistakes.

• Researchers claim breakthrough in fight against AI’s frustrating security hole

  Prompt injections are the Achilles' heel of AI assistants. Google offers a potential fix.

• 4chan has been down since Monday night after “pretty comprehensive own”

  Early info is unreliable, but the site has been mostly unavailable for hours.

• OpenAI continues naming chaos despite CEO acknowledging the habit

  OpenAI's brand-new "GPT-4.1" has a funky name but reasonable performance for the price.

• Amid Trump tariff chaos, Nvidia launches AI chip production on US soil

  New announcement seems aimed at appeasing Trump, signals company shift.

• That groan you hear is users’ reaction to Recall going back into Windows

  Snapshotting and AI processing a screen every 3 seconds. What could possibly go wrong?

• Researcher uncovers dozens of sketchy Chrome extensions with 4 million installs

  Even weirder: Why would Google give so many the "Featured" stamp for trustworthiness?

• Researchers concerned to find AI models misrepresenting their “reasoning” processes

  New Anthropic research shows AI models often fail to disclose reasoning shortcuts.

• OpenAI helps spammers plaster 80,000 sites with messages that bypassed filters

  Company didn't notice its chatbot was being abused for (at least) 4 months.

• After months of user complaints, Anthropic debuts new $200/month AI plan

  Two-tiered "Claude Max" expands rate limits and offers traffic priority to subscribers.

• “The girl should be calling men.” Leak exposes Black Basta’s influence tactics.

  Disclosure of tactics, techniques, and procedures provides rare glimpse into secretive group.

• Carmack defends AI tools after Quake fan calls Microsoft AI demo “disgusting”

  Carmack: "Power tools" like AI are "central to all the progress in computers."

• Meta’s surprise Llama 4 drop exposes the gap between AI ambition and reality

  Touted 10M token context proves elusive, while early performance tests disappoint experts.

• NSA warns “fast flux” threatens national security. What is fast flux anyway?

  Used by nation-states and crime groups, fast flux bypasses many common defenses.