Google wins by losing
Welcome to Cautious Optimism, a newsletter on tech, business, and power.
Wednesday! Google stock is up sharply today after remedies for its search monopoly trial came in less gory than feared (see below). Elsewhere, US government debt continues to sell off, the Fed is fighting political interference, and we have so much to talk about. To work! — Alex
📈 Trending Up: Tariff refunds? … Happy Robot … my age viz the news … Waymo’s geographical footprint … US and Indian VC harmony … an ‘inflation mountain?’ …
📉 Trending Down: UK manufacturing … free speech … stablecoin doubts … IRA grants … inflation in Ghana … inflation in South Korea …
Things That Matter
Tariffs have real impacts: While tariffs have proved moderately inflationary thus far, we’re starting to see the impact of rising imports on business sentiment. The Dallas Fed polled Texas business leaders, showing that 48% of “businesses said they’ve been negatively impacted by higher tariffs this year,” with more than 70% of manufacturing businesses reporting negative business impacts, and some 60% of retail firms.
The survey also reports that amongst the Texas businesses polled, just under “half of firms reported increased input costs as a result of higher tariffs,” leading to 40% of firms reporting margin compression. So, what’s ahead? Higher prices, with 48% of firms “negatively impacted by tariffs” saying that they are passing on “cost increases to customers.”
Apple’s AI posture: Apple’s work to release AI models that are tuned for on-device inference is good fun, even attracting public plaudits from tech leaders. But while Apple’s AI story might be a bit undersold in the face of criticism regarding the slow rollout of Apple Intelligence, the company is shedding staff at a critical rate. I am not saying that Apple needs to go dollar-for-dollar with Meta in the AI recruiting game, but it does need to do something to plug its intra-Silicon Valley brain drain.
AI demand bolsters Zscaler: Cybersecurity firm Zscaler reported in its earnings call yesterday (transcript) that it is seeing “seeing significant customer interest in our powerful AI security solutions, including our new AI Guard and GenAI Security offerings.” Why? Per the company, the market has “entered an era of omnipresent AI,” which is driving “adoption of AI at this breakneck pace,” leading to “new security challenges such as model jailbreaking, prompt injection, model poisoning and more.”
Many AI companies are trying to make money off models, infra, or apps. There are other good ways. Astera Labs sells data center connectivity tools, and has performed brilliantly since going public. Zscaler is minting growth on the back of keeping AI secure for business customers. Lots of rising boats in the AI era, it seems.
Anthropic lands the plane: It was nice to see Anthropic confirm its $13 billion round valuing the comapny at $183 billion (post-money) yesterday. But more importantly, the company underlined reporting that it grew from $1 billion ARR to $5 billion from ~January to ~August, while growing its number of $100,000+ accounts 7x in a year. Claude Code is also at a $500 million run rate already. The folks who can see the underlying financial data and future roadmap are bullish. The rest of us? Sitting in a post-GPT-5 funk, I suppose.
OpenAI buys Statsig: Here’s a weird one. OpenAI has purchased Statsig, making its founder and CEO Vijaye Raji CTO of Applications under Fidji Simo, the AI model giant’s CEO of Applications. What matters is that Statsig was valued at $1.1 billion earlier this year on the back of ~$40 million worth of annual recurring revenue. And it is selling to OpenAI for the same price tag. What gives? Preusmably Statsig backers were content to alchemize their Statsig stock into a blend of cash and OpenAI stock, with earlier backers of the startup seeing reasonable profit on their investment.
Conflict alert! Statsig’s last round was led by ICONIQ. The same investing firm that just co-led Anthropic’s Series F. Congrats on playing both sides, ICONIQ.
Google wins by losing
After Google was deemed to hold a monopoly in the Internet search market, a slate of possible remedies was floated. These included a ban on paid search deals, an end to exclusive search deals with companies like Apple, the divestment of Chrome, and Mountain View being forced to share some of its search data to foment competition.
You can read the ruling here. A good overview of industry complaints concerning whether or not the penalties go far enough can be found here.
The judge overseeing the case — U.S. District Judge Amit P. Mehta — decided that:
“Google will be barred from entering or maintaining any exclusive contract
relating to the distribution of Google Search, Chrome, Google Assistant,
and the Gemini app,” along with rules that preclude it from leveraging its Android operating system to prejudice the market towards its search products.
“Google will not be required to divest Chrome […] nor will the court include a
contingent divestiture of the Android operating system in the final
judgment,” a huge win for the search giant.
“Google will not be barred from making payments or offering other
consideration to distribution partners for preloading or placement of Google
Search, Chrome, or its GenAI products,” which means Apple can continue to hoover up huge sums from the search company.
Google will share data with competitors to break the cycle of its having more users, and thus more data, and thus better products, and thus more users, and thus — you get the idea.
And Google won’t have to offer a search ballot screen, become “subject to an investment reporting requirement,” or “modify its policies to offer website publishers more choice in how Google uses their content.”
Why did the judge pull their punches at the end of all that legal wrangling? Because genAI used in a search context is attracting enough capital and sufficient search volume to constitute a competitor to Google’s “general search engine.” Here’s the key quote (emphasis added):