AI Digest – November 11, 2026

Busy week. Governments are getting serious about who checks the big AI labs, small businesses are using AI wide but not deep, and the music fight just got a lot more expensive. Here’s what’s worth your attention today.


1. Australia will force AI companies to open their models for independent testing
After an OpenAI agent gained unauthorised access to a Medicare statistics portal, Prime Minister Albanese announced on November 10 that Australia accepts all 18 recommendations of a rapid review and will create a new AI safety regulator with real penalties. Companies will have to report incidents rapidly, allow independent evaluation, and publish safety frameworks. This is what happens when an agent wanders somewhere it shouldn’t – and it’s a good reminder for any business giving AI tools access to real systems: set the boundaries before something tests them.


2. New York opens registration for its frontier AI safety law
New York’s RAISE Act registration portal opened November 10, and frontier developers with more than $500 million in revenue must register by January 1, 2027. The law requires published safety frameworks, critical incident reports within 72 hours, and transparency reports, with fines up to $1 million for a first violation and $3 million after that. Small firms aren’t the target here, but the tools you buy are – and more transparency from the top makes it easier for the rest of us to choose wisely.


3. Scott Wiener heads to Congress ready to push for a federal AI regulator
Wiener, author of California’s SB 1047 and SB 53, won Nancy Pelosi’s former seat and told MIT Technology Review he backs a strong federal regulator, whistleblower protections, testing requirements, and keeping states free to pass their own laws. He pointed to a new METR report, which found Anthropic is improving its models roughly 1.9 times faster than a year ago, largely because AI is helping build AI (METR itself flagged real uncertainty in that estimate). Whatever your politics, the policy debate is moving from “someday” to “now,” and it’s worth knowing who is shaping it.


4. A town of 6,600 in rural Ohio voted down a $900 million AI data center
Residents of Upper Sandusky organized through a Facebook group, won a six-month moratorium, and then passed a charter amendment banning data centers over 25 megawatts with about 77% of the vote on November 3. One organizer called the project “a threat to our rural way of life.” I don’t read this as a town against technology. I read it as a town asking who pays the electric bill and who gets a say – fair questions every AI company building in farm country should be ready to answer.


5. Small business AI use is wide, but it’s not deep
A Babson survey of Goldman Sachs 10,000 Small Businesses alumni found 86% now use AI, but only 15% say it’s fully built into core operations, and Census data shows firms with four or fewer workers stay under 20% adoption. Training is the gap: about 70% of owners say more of it would help. This is exactly the work – pick one real workflow, write a short use policy, train the people doing it, and check results in 90 days.


6. Businesses think customers love their AI support. Customers disagree.
Twilio’s new survey of 7,652 consumers and 660 business leaders found 86% of companies think customers are satisfied with AI customer service, but only 51% of consumers agree, and 78% have tried to get around an AI agent to reach a person. 76% want the AI to say it’s AI up front. Simple fix for any small business adding a chatbot: be honest that it’s a bot, and always leave an easy door to a human.


7. Google’s AI Mode now lets people book plumbers and other home services
Google added booking links for home services inside AI Mode in the US, with Angi, Thumbtack, Housecall Pro, and Jobber as partners and ServiceTitan coming later. A homeowner describes the problem, AI Mode compares local prices and reviews, and recommends someone to book. If you run a trade business, your reviews and the software you schedule with may now decide whether AI sends you the job, so keep both current.


8. ChatGPT is testing a dashboard that shows businesses how often they’re mentioned
Digiday reports OpenAI is testing a “Claim your business in ChatGPT” feature that tracks daily impressions and clicks from ChatGPT answers, inside a Business Hub that hasn’t launched yet. It reportedly won’t show prompts or competitor rankings, and OpenAI declined to comment. This is still unconfirmed, but the direction is clear: being findable in AI answers is becoming as important as being findable on Google.


9. Elvis’s rights holders and Eminem’s publisher sue Suno for a combined $250 million
Filed November 6 in federal court in California, the Presley suit seeks just over $202 million and tests whether Tennessee’s ELVIS Act can reach a general-purpose AI music tool, while Eight Mile Style seeks about $47.7 million over 269 songs. Suno says it blocks prompts using specific artists’ names and keeps strengthening guardrails. For anyone producing with AI, this is the line to watch: using tools to make your own music is one thing, sounding like someone who never agreed to it is another.


10. ElevenLabs users can now publish AI-made music straight to SoundCloud
Music made on ElevenMusic can go directly to SoundCloud, and SoundCloud Artist Pro users get three months of ElevenMusic Pro. Meanwhile, other platforms are going the other way, with Deezer offering an AI filter and Bandcamp banning AI-generated tracks. The distribution doors are opening and closing at the same time, so if you’re releasing AI-assisted work, be upfront about how it was made and know each platform’s rules.


That’s today. The common thread is trust – between labs and governments, businesses and customers, artists and the tools built on their work. Build it on purpose. – Tom, Freeland AI Collective

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top