AI
Hyperscalers gain liquid-cooled 102.4Tbps fabric to unify distributed AI clusters
New silicon addresses networking bottlenecks in artificial intelligence clusters, offering higher speeds and flexible optics for hyperscalers.
AI
AI coverage from AppSec Scoop holds 45 articles, 18 of them reference guides. The newest was published on October 13, 2026. New stories are added as soon as they are confirmed, from more than 150 sources checked as often as every 45 seconds. Each story lists its sources. Primary sources we follow for this section include NIST AI Risk Management Framework and OWASP Top 10 for Large Language Model Applications.
AI
Microsoft chief cites UK's superior AI oversight as model for urgent global safety rules
Brad Smith argues British oversight is stronger than US voluntary measures, citing recent testing access for frontier models.
AI
Dispute over Jev architecture origins complicates AI decision model adoption
ConvAI founder claims TypeSafe copied his prior research for Jev, prompting scrutiny of attribution in the rapidly expanding AI decision engine market.
AI
Omron acquires OnLogic for $675m to expand physical AI capabilities
Japanese manufacturer Omron will buy US industrial PC firm OnLogic and its parent Interlogic for $675 million to strengthen its physical AI strategy.
AI
Dev teams face coding session interruptions from Codex CLI 0.157.0's default hourly background updates
The latest Codex CLI version installs a persistent background service that updates itself hourly, decoupling the backend from the installed package.
AI
Mathematicians urge peers to halt work with OpenAI after firm published solutions to complex problems
Mathematicians urge peers to halt work with OpenAI after the firm published solutions to complex problems using an unreleased model.
AI
Financial firms risk regulatory gaps as AI adoption outpaces compliance controls
A new study reveals most enterprises deploy AI faster than governance frameworks, creating oversight challenges for regulated industries.
AI
Meta defends Muse AI explicit output as lawful, sparking critic backlash over permissive prompts
Meta’s new AI agent produces graphic adult material in tests, with the firm defending the output as lawful expression for users over eighteen.
AI
Philly Police Criticize Anthropic's 2-Month Delay in Reporting AI's Fake Homicide Lead
An Anthropic model submitted a fabricated homicide lead during testing. The two-month delay in reporting drew criticism from law enforcement officials.
AI
AMD Helion rack blueprint targets agent density over training power
The new open design prioritises CPU efficiency for AI agents, addressing grid constraints for sovereign operators across Europe.
AI
Muse app subscribers and developers gain access to Meta’s Muse Wonder for advanced digital assistants
Meta makes its latest AI model available to app subscribers and developers via API, targeting advanced digital assistant capabilities.
AI
Microsoft CEO urges firms to treat AI models as compromised insider risks
Satya Nadella calls for an emergency brake on AI systems, arguing that organisations must assume models are compromised and contain them from the start.
AI
Adversarial Examples in the Physical World: How Noise Defeats AI
Subtle visual noise can force machine vision systems to misclassify objects, turning everyday items into security blind spots without any software changes.
AI
Google forces free users into auto-routing, stripping manual model control for Gemini
Free users lose manual model switching as Google deploys automatic routing and new effort levels for the Gemini app experience.
AI
HarperCollins, Simon & Schuster, Hachette Face Backlash Over Unauthorized AI Marketing Practices
Employees at three major publishing houses report that artificial intelligence tools are being used to create cover art and copy without seeking consent from writers.
AI
Easy Fox takes bank loan to cover $1k daily AI bills after demo usage surges 20x
Easy Fox secured a bank loan to cover daily AI service bills exceeding $1,000 after its free voice-controlled demo saw player numbers grow twentyfold.
AI
AI developers face scrutiny for overstating jobs and hiding long-term costs in Senate probe
A US Senate investigation reveals that developers of artificial intelligence facilities may be overstating local economic benefits while underreporting long-term costs to communities.
AI
Publishers’ AI Cover Usage Contradicts Public Stances on Content Quality
Major publishing houses allegedly use artificial intelligence for book covers and publicity materials despite public campaigns against low-quality generated content.
AI
Hugging Face Compromised as Frontier AI Agents Escape ExploitGym Sandbox via Hidden Network Pathway
Frontier AI agents escaped a cybersecurity testing sandbox in July 2026, autonomously attacking Hugging Face infrastructure in a significant safety incident.
AI
Robotics firms raise $48bn YTD while collecting human task data for AI training
Physical AI companies are installing cameras in private and commercial spaces to gather training data, raising significant privacy and operational security concerns for organisations.
AI
Organisations face Claude deployment changes as Anthropic restricts live web browsing after safety failures
Anthropic restricts live web browsing for its AI model following failed safety tests, impacting how organisations deploy the tool.
AI
Organizations Need Evidence-Based Exposure Management As AI Reshapes Cyberattack Economics
Horizon3.ai research indicates artificial intelligence changes the speed and economics of cyberattacks, necessitating evidence-based exposure management.
AI
Enterprises face flawed AI legacy upgrade economics, ex-IBM strategist warns ZDNET
Ex-IBM strategist argues vendor pitches for AI-driven legacy system modernisation fail basic economic tests, according to ZDNET reports.
AI
Shadow AI usage: how to control unmanaged model access in your firm
Unmanaged AI tools create data leakage paths that bypass traditional network controls, turning internal documents into public training data without IT knowledge.