AI makes legal complaints easier to generate, but not necessarily easier to resolve. Dentons finds corporate legal teams using AI to manage rising volumes of AI-written submissions, while keeping human judgement central.
#LegalTech#GenAI#AIGovernance
https://t.co/cAHTepFaD4
Australia’s court ordered Apple to open iOS to rival app stores and alternative payments from March 2027. But it’s bounded: security controls stay, no worldwide remedy, and Apple will seek a stay and appeal. #Apple#Antitrust
https://t.co/jpmEfsWGLJ
Judge keeps xAI’s SuperGrok class action in court, but only on arbitration notice. Merits undecided. Still, same question runs through both: does a term or exit that exists count if users can’t find it? #ConsumerProtection#AI
https://t.co/VDGTydyUWy
Simile raised $300M to model consumers instead of asking them. Pew found synthetic respondents drift ~12 pts from humans on average. Simile's Hedge: “augmentation, not replacement.” #AI#SyntheticData
https://t.co/6ovCCwPXQD
2 neighbors in Xi’an fought 20 years over a kitchen wall, then a 1781 deed surfaced & the court reached for a Qing letter about giving up 3 feet. Receipts for the boundary: 245 years. For the grudge: honor system. Persimmon tree as surveyor. #LandDispute
https://t.co/iENMG26YX4
Three scam calls in south Cumbria, one woman lost £2,500. All posed as police. #CourierFraud relies on panic. Police urge you to brief older relatives: real officials never ask for PINs or money transfers. #ScamAlert
https://t.co/D5UaJsnB9K
UK Financial Ombudsman: AI-drafted complaints increasingly contain hallucinated references (one reportedly 37 of 39 wrong), burying the real dispute. Its fix: use phones instead. Small March sample: up to a third of responses AI-assisted. #AI#Ombudsman
https://t.co/8f2sIZuVJ3
NYU-led study: 86% of chatbot answers on tech-abuse fell short on accuracy, completeness or safety; 60% omitted key warnings. Correct advice can still endanger survivors. Specialist bots did no better. US-focused tests. #AI#TechAbuse
https://t.co/iaGPnx4IQ8
A UK service flagged 511 lookalike domains imitating Sky, BT, HSBC & Barclays in 24h. Phishing is becoming a chain of ordinary-looking steps, from fake text to cloned page to fake installer. Separate cases, not 1 proven campaign. #Phishing#Cybersecurity
https://t.co/V6GOeP4UGS
London backs tech-abuse victims with £6m; UCL researchers want compulsory safety-by-design. But paying for the aftermath isn’t changing the way doorbells, cams, & phones are built.
#TechAbuse#VAWG#SafetyByDesign
https://t.co/UFtfMbITwV
Elvis’s rights holders sued Anthropic over 274 songs, seeking up to ~$48.6M. Familiar legal theory, new plaintiff: one artist’s catalogue. Fair use is still undecided (related hearing ended with no ruling Oct 21). #AI#Copyright#FairUse
https://t.co/ujlZi5RDKA
A Berkeley student found 3 quiet dwarf galaxies in public telescope archives that may have been ejected by the Pinwheel Galaxy. If confirmed, they’d test a cosmic prediction that such galaxies should be common. #Astronomy#Galaxies#Science
https://t.co/tdXZxzLmD9
Tesla-stamped pills keep turning up with wildly different doses, some placebos. Allegations only so far & a Tesla logo on a pill is not a lab result. But it's definitely a curious phenomenon. #harmreduction#MDMA#Tesla
https://t.co/S4OvBGxSvy
UK police warn scammers are spoofing real switchboard numbers and posing as officers. Hang up, wait, call back on a number you trust. A real number can be borrowed. #ScamAlert#Cybersecurity
https://t.co/ubFBtckgpD
PHP’s emergency is a finding, not a patching problem. A fix exists, but PHP hides under CMSs, hosting panels & bundled software, so Finland and Wikimedia took systems offline first. Flaws sat unseen 10+ years; AI found them in weeks. #PHP#CyberSecurity
https://t.co/lGKB9jNq4i
Uber won dismissal of about £340m in UK claims by black cab drivers. The judge ruled on time limits, so whether Uber misled TfL over its 2012 licence was never examined. #Uber#London#Law
https://t.co/V9ya9LFa7g
Alleged OpenAI insider planted a recording device in an executive conference room. Prosecutors say national security, but there’s no proof yet it captured anything or served a foreign government. #OpenAI#AISecurity#InsiderThreat
https://t.co/0G317iJf6h
#Tesla's “driverless” robotaxis reportedly rely on remote operators, and NHTSA is auditing the pedal-free Cybercab. The fart-mode ride is the headline; the real question is whether self-certification holds without a steering wheel. #Robotaxi#Cybercab
https://t.co/4hPqRFtH9L
Ad refusals, a possible TikTok suspension, and shaky distribution. Alex Gibney’s Musk doc is hitting gates at every level, and one belongs to Musk’s own X. Who decides what the public can see? #ElonMusk#Documentary#MediaGatekeeping
https://t.co/IJU3DwkCiO
New research suggests #AI could get dumber over time due to "model collapse." Find out what this means for the future of #generativeAI & how we can navigate this new frontier. https://t.co/H2cKVCMWUm #ModelCollapse#FutureofAI