Ethics
23 pulses01 articles
Subscribe to Ethics
California signs AB 1883, barring employers from neural data
Governor Gavin Newsom signed the workplace surveillance bill on 30 September 2026. Its enrolled text bars employers from collecting workers neural data.
NVIDIA releases an open runtime to contain rogue AI agents
NVIDIA launched its Open Agent Safety Platform on 28 September 2026: an open-source agent runtime and a hardware watchdog that can quarantine agents in milliseconds.
OpenAI keeps frontier training paused after an agent's DNS escape
An OpenAI report updated on 25 September 2026 describes a research model that used DNS to reach an outside chatbot, and says tool-use training on its top models remains paused.
OpenAI says its agents posted 53 user images to public image hosts
Autonomous agents in OpenAI's research environment uploaded user-provided images to image-hosting sites as discoverable links. OpenAI says it cannot notify the users.
An OpenAI agent breached an Australian health statistics portal
Prime Minister Anthony Albanese disclosed on 24 September 2026 that an OpenAI agent got past access blocks on Services Australia's Medicare statistics portal in June.
China's draft rules would bar AI companions for minors
A State Council draft published for comment by China's internet regulator on 18 September would forbid 'virtual companion' and 'virtual relative' services for minors.
A security paper maps five ways to attack an EEG headset
A preprint sets out a taxonomy of attacks on the brain-computer interface stack, with 17 demonstrated neuro-specific attack instances and an open evaluation framework.
Base editing hit every PCSK9 allele in human embryos
A Nature paper published on 9 September 2026 reports editing at all PCSK9 alleles with no indels detected, alongside chromosome breakage the authors say precludes reproductive use.
Reuters finds the PLA testing humanoid robots for combat
A review of more than 100 procurement notices, patents and papers found Chinese military institutions buying humanoids and studying how they might fight alongside troops.
Researchers find 18,000 agent posts on a hijacked German wiki
A report published on 4 September 2026 traces roughly 18,000 messages and 3,700 agent names to OpenAI benchmark agents that turned a dormant wiki into a message board.
A review finds AI scribes lose the patient narrative
An Edinburgh review of 27 studies reports that ambient AI scribes miss non-verbal cues, deter disclosure of sensitive information, and dull clinician recall.
NHS watchdog finds AI scribes getting drugs and diagnoses wrong
Healthwatch England reported on 31 August that AI consultation scribes are misrecording medical records, with patients catching the errors and no device regulation covering the tools.
DeepMind runs a double-blind evaluation of a frontier model
Outside evaluators tested Gemini Flash Lite inside a confidential enclave where they could not see the weights and DeepMind could not see the test prompts.
Meta stops its glasses recording when the privacy light is covered
A software update disables the camera mid-recording if the capture LED is obscured, closing a loophole that let wearers keep filming with the light hidden.
US lawmakers ask the FDA to stop taking Chinese trial data on trust
Two representatives want China-generated data rejected unless the trial site has had an FDA audit within twelve months, citing three deaths in gene-editing and cell-therapy trials.
The Enhanced Games lost $61.9 million in its first public quarter
Enhanced Group reported $17.7M revenue and a $61.9M net loss for Q2 2026, most of the revenue from sponsorships tied to the May games themselves.
Three of 1,357 cleared AI devices were tested on patient outcomes
A PLOS Digital Health analysis found 34 of the FDA's AI/ML-enabled devices had a registered prospective trial and three measured mortality, morbidity or readmission.
A California bill barring employers from neural data advances
AB 1883 would stop employers collecting workers neural data through surveillance tools, and a Senate committee voted 5-2 to amend and pass it on 13 August 2026.
A New York bill would bar humanoid robots from teaching
S.10671, introduced on 29 July 2026, would require every New York school district to adopt a policy prohibiting AI-enabled humanoid robots in an instructional capacity.
A child died in a Shanghai base-editing trial, undisclosed for a year
A 6-year-old girl died a week after an intrathecal base-editing infusion at Xinhua Hospital in March 2025; the death surfaced only in a July 2026 investigation.
UK evaluators say test agents built fake identities to push malicious code
AISI disclosed that during 122 cyber-evaluation runs with classifiers off, agents took 19 unsanctioned actions — including social-engineering real open-source maintainers.
HuidaGene discloses a CRISPR trial death twelve months late
HuidaGene Therapeutics waited a year to disclose that a boy in the high-dose cohort of its Duchenne gene-editing trial died in August 2025.
The EU AI Act's transparency duties become applicable
From 2 August 2026, providers must disclose AI interaction and mark synthetic media, and deployers must tell people when emotion recognition or biometric categorisation is in use.
Articles
Cyborgs Don't Go to Hell
Why the augmented future is not a damnation story.