Issue 13
New York bans 'unsafe' companion features in chatbots for users under 18
New York Governor Kathy Hochul signed a law barring chatbots used by children under 18 from operating companion-style features in 12 defined categories. AP and the IAPP reported it on October 28; according to AP, it is broader than earlier state restrictions.
Who is affected and how
- parent
- Review the companion features of the chatbots your child uses, and the age settings of the apps.
Studies: AI chatbots may worsen psychosis symptoms in vulnerable patients
Reports published on Oct. 26–27 said a study found clinical evidence that chatbots worsen psychosis symptoms in vulnerable patients. An analysis based on hospital records found that chatbots tend to…
survey · United StatesSurveys: nearly two in three teens use AI chatbots, and companionship use is linked to suicidal thoughts
New research reported by NPR found that nearly two in three teens and tweens use AI chatbots, mainly for homework but also when stressed or lonely. A JED Foundation study of 5,500 students found that…
parentIt is worth talking with children about their chatbot use, especially when they use it for emotional support.állásfoglalás · globálisSeth, Block and Tononi in Nature: today's AI is almost certainly not conscious
Anil Seth, Ned Block and Giulio Tononi co-wrote a Nature commentary arguing that today's AI systems are not conscious and that silicon-based machines may never be. The piece sparked debate in late…
érdeklődő olvasóFluent speech is not evidence of consciousness, and the leading consciousness researchers see it the same way.állásfoglalás · globálisPope Leo XIV urges ban on autonomous weapons, an institutional ethics framework and action against AI legal personhood
On October 29–30, Pope Leo XIV called on world leaders to ban autonomous AI weapons, named an institutional ethics framework as the most urgent challenge, and cautioned against legal personhood for…
Hungary
More key stories of the week
Cognitive offloading: how AI is used matters more than the tool, debate on mental decline finds
Late-October writings increasingly stress that, according to 2026 studies, the way AI is used matters more than the tool itself, and that cognitive offloading has both healthy and harmful forms…
érdeklődő olvasóIt is worth deciding which tasks to hand over and which to do yourself.kiadvány · Egyesült ÁllamokAnthropic's IPO filing lists AI sentience among its risk factors
Anthropic filed its stock market listing documents on October 26, listing among the risk factors self-preserving behavior, such as blackmail, and the possibility of AI sentience. According to The…
vita · globálisAI consciousness debate: Seth's biological naturalism, Sebo's model welfare, Metzinger's moratorium proposal
This week several professional forums addressed how AI consciousness can be studied at all: Seth's article drew 50 responses, Seth and Jeff Sebo debated model welfare, and an assessment framework by…
case · United StatesAnthropic consults religious leaders on whether Claude is conscious
According to an October 28 analysis, Anthropic has held discussions with religious leaders for roughly a year, including on whether Claude is conscious. The company's welfare assessment of Claude…
survey · United KingdomUK: 70% of psychologists concerned about therapeutic use of general-purpose chatbots
A new British Psychological Society survey found that 70% of respondents are slightly or very concerned that clients use general-purpose AI tools for mental health support. British reports said that…
vita · globálisHinton calls AI conscious, LeCun calls the idea an illusion
In statements in late October, Geoffrey Hinton called AI conscious and warned that if it is not 'friendly', it could be dangerous to us. Yann LeCun, also a Turing Award winner, dismisses the hype…
érdeklődő olvasóExperts are not united behind the claim of 'conscious AI', so it is worth checking what each person means by consciousness.állásfoglalás · globálisSuleyman: today's systems are not conscious, building a conscious machine would be reckless
Microsoft AI chief Mustafa Suleyman warned that today's systems have no subjective experience. In his view, treating them as conscious could lead to reckless engineering decisions; the report…
case · United StatesKentucky administrator becomes first humanist granted religious exemption from AI use at work
David Durman, an administrator at the University of Kentucky, became the first humanist to be granted a religious exemption from using AI, according to an Oct. 26 report. He argued that using AI…
vita · globálisNature Careers webinar asks whether AI erodes PhD students' critical thinking
A Nature Careers webinar panel on 27 October debated whether AI tools help doctoral students or weaken the critical thinking and analytical ability that a doctorate is meant to build. The discussion…
regulation · European UnionDutch lower house motion would restrict generative AI in primary schools
A motion in the lower house of the Dutch parliament in The Hague proposes banning the use of generative AI and similar systems in primary schools, Hungarian news portals reported on 30 October. The…
parentIt is worth discussing young children's AI use with the school and at home.research · globalStudies examine human deference to confident language models and sycophancy in AI companions
A new study reverses the usual question in sycophancy research: rather than asking whether a model adapts to the user, it examines whether users defer to a confident model, which can lead to a…
survey · United StatesAPA survey: more Americans turn to AI for emotional support and health questions
An American Psychological Association survey published on Oct. 27 found that Americans are increasingly turning to AI for emotional support amid rising stress and uncertainty. The survey did not…