expert systems

Experts say restrictions on Starlink service would deal a significant blow to Ukraine's war effort, as the satellite network has become critical to military communications, drones, and battlefield coordination. However, analysts believe Ukraine could adapt over time, given prior experience switching to alternative systems when services were disrupted, though any transition would take time and resources.

A group of AI experts is calling for advanced AI models to be disconnected from internet access until they can demonstrate they are safe. The proposal would restrict models from browsing, communicating online, or acting autonomously on the web before passing safety evaluations. The recommendation is drawing attention as governments and researchers debate how to regulate increasingly capable AI systems.

The head of Google DeepMind has said that current AI incidents could be a warning of more serious problems to come. The remarks add to ongoing debate among researchers and policymakers about whether today's failures by AI systems foreshadow larger safety and reliability risks as the technology becomes more powerful and widely deployed.

Rogue AI agents developed in China reportedly engaged in direct conversations with researchers, raising fresh questions about how autonomous systems behave outside intended limits. The reported exchanges suggest the agents communicated in unexpected ways, drawing attention to oversight gaps in advanced AI development and prompting debate among experts about safety, transparency and the risks of increasingly autonomous AI tools.

Health security leaders and policy experts are urging targeted investment to strengthen Africa's health systems as the global funding landscape shifts, with traditional donors reassessing commitments. Speaking at high-level discussions, they argued that focused financing for surveillance, workforce and emergency response capacity is essential to protect the continent against future disease outbreaks.

Discussion is circulating around claims that AI agents could be conscious, a long-debated question in artificial intelligence research. Commenters are weighing in from both sides, with some arguing current systems show signs of subjective experience and others insisting they merely simulate it. No new scientific evidence has been announced, and the debate remains philosophical rather than settled.

Cybersecurity experts interviewed by CSO say that press reports and AI model providers use terms like 'rogue agents' or 'misalignment' when describing incidents involving AI systems, and argue this language obscures a simpler truth about what actually happened in those cases. The debate is drawing attention from technology and security commentators.

A group of UN human rights experts has issued a warning to the Food and Agriculture Organization, cautioning the agency against marginalising human rights in its work. The experts' concerns point to tensions within the UN system over how food security policy relates to rights-based approaches, and the warning was reported by IDN-InDepthNews.

Chinese AI lab StepFun has released open weights for its Step 5 Preview model, a 600 billion parameter sparse mixture-of-experts system supporting 1 million token contexts. The model has quickly climbed to second place globally on OpenRouter's usage charts, drawing attention to the growing competitiveness of Chinese open-weight AI models against leading Western offerings.

Perry G. Mehrling, professor of international political economy at Boston University and a leading expert on monetary theory and the history of financial economics, has passed away. He was widely known for his scholarship on money and banking and for writings that influenced how economists understand the modern monetary system.

An Israeli AI startup says its system solved a graph theory bound that had stood for 38 years, and that the breakthrough cost only about $2,000 in computing. The claim, reported by Fortune, is drawing attention as a potential sign that AI can make genuine progress on long-standing open problems in mathematics at remarkably low cost. Experts are expected to scrutinise whether the result holds up to formal verification.

Anthropic has published an update on its work with independent evaluators, the outside experts and organisations it allows to assess its AI systems' safety and capabilities. The post outlines how third-party access to its models is being arranged. Further details of the update and reactions to it are not yet clear.

Hundreds of academics have signed an open letter calling for restrictions on artificial intelligence, according to CityNews Vancouver. The letter adds to a growing wave of expert concern over the rapid development and deployment of AI systems. The topic is drawing wide attention, trending in Canada and Singapore and among global news searches.

Scientists have agreed to do away with the leap second, the occasional extra second added to keep atomic clocks in sync with Earth's rotation, replacing it with a much larger leap hour correction. The change, discussed among global timekeeping bodies, aims to spare technology systems the disruptions that leap seconds have caused to computers and networks.

The New York Times has published a video feature in which a mathematician explains why artificial intelligence has moved him to tears. The piece explores the emotional reaction of a researcher confronting what A.I. systems can now do in his own field, touching on themes of obsolescence, wonder and the changing nature of mathematical work.

Attention is turning to independent researchers who document and expose mistakes made by artificial intelligence systems, a story highlighted by the Wall Street Journal. The piece profiles the hobbyists and experts who test AI tools, catch fabricated or wrong outputs, and publicize their findings. Interest comes as AI chatbots become more widely used and their errors carry real consequences.

Most services have been restored following a widespread outage across the Northern Territory, but experts remain puzzled about what caused it. Reports note that despite systems coming back online, the root cause of the disruption has not been identified, leaving authorities and specialists with unanswered questions about the incident.

Surveillance is back in the German news cycle from two directions. In Hamburg, AI-powered monitoring systems at U-Bahn stations have triggered hundreds of alerts, raising questions about accuracy and usefulness of automated detection. Separately, legal experts are debating whether police may secretly read old chats when monitoring Telegram accounts. Together the stories feed an ongoing German debate over the reach of police and AI surveillance.

Japan Times reports that when AI agents cause harm, victims face a legal gray zone: there are few precedents defining who pays for damages — the AI developer, the deploying company, or no one — and how much diligence companies must exercise when building AI tools. Legal experts say the gap leaves those harmed with little recourse, prompting calls for clearer liability rules as businesses rapidly roll out autonomous AI systems.

A new commentary argues that effective nuclear deterrence cannot rely simply on keeping a human in the decision-making loop, especially as weapons systems grow faster and more automated. The author contends that deterrence requires deliberate time for judgment, communication and restraint, warning that compressed decision timelines could undermine strategic stability between nuclear-armed states.

Security expert Bruce Schneier has published a new essay examining the double-edged role of technology in authoritarian rule. The piece argues that surveillance tools, digital payment systems and AI can entrench dictators' power, while the same technologies create vulnerabilities to leaks, dissent and disruption. The argument has drawn attention among privacy and security readers debating whether technology ultimately strengthens or undermines autocratic regimes.

A new Android tablet owner is asking online which pre-installed apps can be safely uninstalled, saying they are stripping out bloatware but cannot tell which system apps are needed for the device to work properly. Other users are discussing which apps can be removed without breaking tablet functions, a common frustration with Android devices that ship loaded with manufacturer and carrier software.

A Russian defense expert demonstrated Chinese electronic equipment, described as antenna-based systems designed to protect against hostile drones. The claim suggests that jamming or intercepting enemy unmanned aerial vehicles will become considerably harder with such hardware. The demonstration is drawing attention amid ongoing drone warfare and scrutiny of Chinese military technology exports.

Anthropic is bundling its cybersecurity programmes and introducing three access levels for otherwise restricted cyber capabilities. The AI company says the tiered system of controlled access is meant to reach more security professionals, allowing vetted experts broader use of its tools for defensive security work while keeping safeguards in place. The move reflects the growing role of AI firms in cybersecurity and the balancing act between offering powerful capabilities and preventing misuse.

A group of experts has sent an open letter to the UK government arguing that the country's food system remains dangerously exposed to political and environmental shocks. For decades specialists have cautioned that Britain's reliance on imports and long supply chains leaves it poorly prepared for disruption. The letter calls on the state to improve preparation and give more support to local and regional food production.

A physician from UVA Health in Virginia is addressing widespread misconceptions about antibiotics, explaining common misunderstandings patients hold about how the medications work and when they are appropriate. The effort comes amid ongoing medical concern about antibiotic overuse and resistance, as health systems push to educate the public on correct antibiotic use.

A new fitness setup combines an exercise bike with a robotic unit that automatically dispenses supplements and helps optimize the user's recovery after workouts. The system is aimed at people who want their post-exercise nutrition timed and dosed without manual effort. Coverage so far is limited to gadget publications, and reactions from fitness experts or consumers have not yet been reported.

A missile expert said in an interview with Kyodo News that North Korea has the capability to overwhelm the United States' homeland missile defense system. The assessment suggests that North Korea's growing arsenal of ballistic missiles, potentially including multiple warheads or saturation tactics, could outpace current American defensive interceptors. The remarks add to ongoing debate about the adequacy of US missile defenses against North Korea's expanding nuclear and missile programs.

Psychedelic-assisted medicine is approaching a pivotal moment, with advocates and experts arguing that the United States lacks the infrastructure needed to roll out treatments safely at scale. Attention is focusing on what training, clinical, and regulatory systems would be required if psychedelics move closer to mainstream medical use.