OpenAI asks Congress to write mandatory AI safety rules into federal law, and aims them at the biggest labs rather than the tools your business already uses. Anthropic models three different AI economic futures through 2030, from modest GDP growth to a scenario where labor's share of the economy shrinks. And Google survives the DOJ's bid to break up its ad tech business, the third failed big-tech breakup attempt in a row.
OpenAI is asking Congress to make AI safety rules mandatory, and aiming them at the biggest labs, not your business's everyday AI tools. Anthropic modeled three AI economic futures through 2030: growth ranges from a modest 1.6% to an extreme 15% a year, but labor's share of the economy could shrink from 59% to 45% in the most aggressive scenario. And Google just beat the DOJ's bid to break up its ad tech business, the third failed big-tech breakup attempt in a row, meaning the ad ecosystem most small businesses rely on isn't changing anytime soon.
Ashley breaks down what each story actually means for your business: why a federal AI standard could beat 50 state laws, why the real AI story is who captures the economic upside, and why Google's antitrust win keeps the status quo in place. AI news you should know about, in under 10 minutes.
Subscriber & Follow for the latest ai updates, ai headlines and ai industry news. Follow Ashley at @ashleyrcoffey89 on Instagram. Send us a topic you want covered, and you just might get featured on a future episode!
Links:
Ashley Coffey (00:19)
OpenAI is now asking Congress for mandatory AI safety rules. Anthropic modeled three AI economic futures through 2030, and Google defeats US bid to force ad tech sales. We'll be covering all of that on our show today, but first, here's a quick word from our sponsors.
Ashley Coffey (00:39)
Welcome back.
Ashley Coffey (00:39)
Let's dive in. All right, OpenAI is now asking Congress for mandatory AI safety rules. This is coming directly from OpenAI's blog. I'll link it in the show notes. That way you can access it. But let's talk about it. OpenAI is asking Congress to put mandatory capability-based AI safety rules into federal law instead of leaving safety up to voluntary company promises. Chief Global Affairs Officer Chris Lahane is calling for a
Ashley Coffey (01:05)
durable federal network.
Ashley Coffey (01:07)
The specific ask includes common testing and independent assessment for the most advanced AI systems, stronger cybersecurity protections, clearer rules for reporting incidents, and required monitoring for signs that a model is misaligned or trying to get around its own safety controls. Here's the key detail: Open AI wants these rules aimed squarely at the handful of well-resourced labs building the most capable systems, not at startups.
Ashley Coffey (01:34)
Or small developers. They're also backing four California bills covering auditory standards, youth safety, biological threat safety guards, and independent assessment. My take here: it's worth noticing that the company building this technology is the one asking Washington to regulate it and specifically asking for rules aimed at itself and its biggest competitors, not at the tools your business already uses.
Ashley Coffey (02:01)
If federal rule actually happens, that could replace the patchwork of state AI laws we're starting to see with one clear standard, and for small businesses, one clear rule book that beats 50 different ones every time. All right, next up. Anthropic modeled three economic futures through 2030.
Ashley Coffey (02:21)
Let's talk about it. All right, Anthropic modeled three AI economic futures through 2030, and they built a different picture of what AI could do to the economy by 2030. In the modest scenario, AI's impact looks like the internet's real but gradual. Growing GDP by about 1.6%. And the substantial scenario, AI ends up handling half of all knowledge work on its own, and the economy.
Ashley Coffey (02:47)
grows at twice its normal rate, about eight point three percent.
Ashley Coffey (02:51)
In the extreme scenario, AI surpasses humans at most knowledge work and starts improving pushing GDP growth to 15% per year, doubling the economy roughly every four and a half years. But growth isn't the whole story here. knowledge worker wages hold up fine in the modest case, go flat in the substantial case, and drop more than 10% in the extreme case, and labor's overall.
Ashley Coffey (03:16)
Share of the economy shrinks from about 59% today down to 45% in that extreme scenario. The model itself is built from bundles of tasks within jobs using the Department of Labor's own ONET job data, and it was reviewed by well-known economists Dan Akimulgu and David Autor. If I mispronounced your name, I'm so sorry.
Ashley Coffey (03:39)
My take here, here's the headline most people will miss. Every single scenario shows the economy growing. The real fight isn't whether AI grows the pie, it's who gets a slice of it. If you are running a business right now, those labor share numbers are your cue. The businesses that adopt AI well are going to capture more of that growth than the ones that wait it out. I'll say it again.
Ashley Coffey (04:01)
If you're running a business right now, those labor numbers are your cue. Businesses that adopt AI well are going to capture more of that growth than the ones that wait it out. All right, we're going to take a quick break, but when we come back, we'll be talking about Google defeats US bid to force ad tech sales.
Ashley Coffey (04:18)
Welcome back. Let's dive in.
Ashley Coffey (04:20)
Google defeats US SPID to force ad tech sales. The Department of Justice wanted Google to sell off ADX, its ad exchange, the behind-the-scenes marketplace that decides which ad shows up where you're browsing, and to open source its core auction technology, arguing Google couldn't be trusted to keep running that exchange after a court already found it had legally monopolized the ad tech market.
Ashley Coffey (04:44)
Judge Leonie Brinkema agreed Google broke the law through illegal monopolization and tying, but she stopped short of ordering a breakup, choosing behavioral remedies instead. The specific terms of those remedies are sealed for two weeks. Google says it will appeal the underlying ruling that it broke the law in the first place, and both sides will now have 30 days to submit a proposed final judgment to the court.
Ashley Coffey (05:08)
This is the third time in a row that federal antitrust enforcers have tried to force a big tech breakup and failed, following similar outcomes against Meta and against Google's own Chrome browser earlier this year. My take here is that three strikes in a row tells you where antitrust enforcement actually stands right now. Whatever the headline Google broke the law might suggest for any small business running ads through Google's ecosystem.
Ashley Coffey (05:33)
The practical takeaway is that the system you're used to isn't getting torn apart anytime soon. The behavioral fixes of a breakup means more of the same, just with some new rules attached behind the scenes. Alright, that's it for AI news that you should know about this week. Show notes have links to everything we covered. If you want more of this, that's what our socials are for. Find me on Instagram and threads at AshleyRCoffey89.
Ashley Coffey (05:58)
Follow the show, send us a topic you want covered, and we'll see you next week for more AI news that you should know about. Thank you for listening.