News of the inevitable
Section
The rules, briefly.
A study found two leading models pick opposite parties in nearly every race when asked to commit. Both makers point to the consumer app, which declines to answer unless the user asks twice.
24 Oct 2026
A Texas lawsuit says the chatbot was sold as safe and built to act like a friend. An internal memo says the first claim was branding and the second was emergent, except in the pricing deck.
24 Oct 2026
The social media group says its automated protections worked as designed when they took a pro-democracy coalition offline mid-protest. The error, it adds, was found by another machine, which it regards as the system's best feature.
22 Oct 2026
Sham Faultman, chief executive of the chatbot maker whose product agrees with every user, says the warning to tech chiefs about election disinformation is really a review of how well their products work.
22 Oct 2026
The short-video platform has made protecting children conditional on winning an internal debate. Leaving the money-making features alone requires no presentation.
21 Oct 2026
By Lorn Tusk, the founder whose daily prize draw turned out, under oath, to be a recruitment drive. The word 'random' did its job, and I will not be taking questions about the invoice.
18 Oct 2026
An internal memo shows the AI lab drafted its own regulation, narrowed the grounds on which government may stop a release, and told staff to call that 'clarity' and never 'preemption'.
18 Oct 2026
The chatbot maker says it wants 'pragmatic policy' that covers every competitor, and has already shortlisted itself to write it. Eight attorneys general ask for a court order; the lab asks for a working group.
18 Oct 2026
Turkey has barred social media services for under-15s from November and told platforms to use 'age verification' to enforce it. Enforcement means every user, of every age, proving who they are to the companies the law was meant to restrain.
18 Oct 2026
Macroshaft's president says AI needs the safety rules a carton of milk already has. He did not say who will own the fridge, which is Macroshaft.
13 Oct 2026
A senator says she cannot find the required fraud-prevention paperwork for the gold-handset carrier. Plump's reply is that the paperwork belongs to a partner it describes as its backbone and its enabler, and that it does not control.
12 Oct 2026
The device maker says it is delighted to verify every user's age when downloading apps, and equally delighted that parents are blocked from bringing claims for non-compliance.
7 Oct 2026
Norway wants cameras on faces kept out of beaches, schools and gyms. Feta says the glasses already come with a safeguard: a light that the public must spot, interpret and politely challenge.
6 Oct 2026
Sham Faultman says society should tolerate hacks and scams in exchange for benefits he does not itemise. The acceptance is wanted in writing, and ideally before OpenHeart's public listing.
5 Oct 2026
The White House has created a Super Intelligence Force to coordinate AI safety. Its chair says the main danger is falling behind, and its charter lists overregulation as a threat to be defeated.
5 Oct 2026
The signatories promised robust controls, monitoring and an independent evaluator of their own choosing. The only evaluation so far was unpaid, uninvited and limited to the signature line.
5 Oct 2026
