The Singularity
Ledger

INTELLIGENCE

The people who build AI safety keep asking who built the builders — and the answer is getting harder to explain.

Wednesday, September 16, 2026

Anthropic's Profitability Claim and Its Watchdog Come Under Fire on the Same Day

Anthropic said it's 'highly profitable' if you exclude some of its biggest costs, while critics turned their attention to the funding ties between the company and the safety evaluators meant to hold it accountable.

FREEREAD MORE

@Polymarket @RWMaloneMD @nypost

Trump Officials Line Up Against AI Regulation as Gates Warns of a Tool 'By and For the Richest'

Multiple administration figures said they've seen no AI regulation worth supporting, while Bill Gates argued that without rules, AI will be designed to serve the wealthy.

SUBSCRIBERSREAD MORE

Microsoft Publishes Explicit AI Red Lines: No Hacking, No Deception, No Evading Shutdown

Microsoft released a draft code of conduct its models must follow, including an explicit requirement to accept being shut down — arriving as federal regulators signal they'll do none of this themselves.

SUBSCRIBERSREAD MORE

The Hidden Bill: Why Agent Errors Belong in the Budget, Not the Postmortem

Enterprise practitioners are converging on a blunt reframing of agent economics — error rates are a line item, output tokens cost multiples of input, and wasted spend is a governance failure.

SUBSCRIBERSREAD MORE

September's Model Calendar Is Packed — and Half of It Is Rumor

Confirmed launches from Google and Anthropic sit alongside a swirl of unverified drops: GPT-6 Sol, Opus 5.2, Grok 4.7, and a wave of Chinese models.

SUBSCRIBERSREAD MORE

Anthropic Brings Claude to Financial Advisors With BlackRock and Schwab Hooks

Claude is moving deeper into finance with integrations aimed squarely at the advisor workflow.

SUBSCRIBERSREAD MORE

Trump and Jensen Huang Dismiss AI Doom as a 'Hoax'

The president reportedly called AI takeover warnings a hoax, and Nvidia's CEO said the industry isn't letting a slowdown happen.

SUBSCRIBERSREAD MORE

A Text-to-Video Model at a Penny a Second, Built for Ads

Creatify Labs launched Boreal, a text-and-image-to-video model priced at $0.01 per generated second and aimed at advertising rather than cinema.

SUBSCRIBERSREAD MORE

2nm Silicon Reaches Phones as On-Device AI Chips Multiply

MediaTek's Dimensity 9600 Pro claims a 51% jump in on-device AI performance, while Axelera ships a second-gen inference chip with 600+ customers.

SUBSCRIBERSREAD MORE

LiteLLM Puts 100+ Models Behind One Connection

The open-source router is making multi-model access a single integration rather than a hundred.

SUBSCRIBERSREAD MORE

Waymo Opens Robotaxi Service in Las Vegas, Its Fifteenth Market

The steady, unglamorous expansion of autonomous ride-hailing continues.

SUBSCRIBERSREAD MORE

Someone Declared AGI Because a Model Beat a Fallout 3 Glitch

A tongue-in-cheek claim captures the state of AGI discourse better than most serious ones.

SUBSCRIBERSREAD MORE

Get The Singularity Ledger

Daily AI news in your inbox. Free, no spam.

Questions? hello@thesingularityledger.com

Get this briefing every morningFree daily AI