AI Stocks & Markets

Today's Stories

1. 1

   Offline smartphone AI accurately detects DR, glaucoma, and AMD in real-world study

   Offline smartphone AI accurately detects DR, glaucoma, and AMD in real-world study

2. 2

   AI Investment Peak Risk: Mag-7 Stocks Down 12.2% From May High

   Bloomberg's Mag-7 index (tracking major AI-focused tech firms) is now 12.2% below its May 29 peak. The comparison to the dot-com boom raises a key question: if equity prices have peaked, will business investment in AI equipment and software follow the same downturn pattern that occurred in 2000–2001. Unlike the dot-com era, when equity declines directly choked off investment funding, today's Mag-7 firms have sufficient retained earnings to internally fund spending and are only now tapping debt markets. This structural difference suggests AI investment may not crater as sharply as it did two decades ago, even if stock prices remain under pressure.

   Professional forecasters assume continued investment growth, but the recent equity decline suggests a possible downturn may be imminent. The actual trajectory of nonresidential fixed investment in equipment and software over the next few quarters will signal whether the 2000 analogy holds or breaks.

3. 3

   OpenAI, Anthropic to go public in late 2026–early 2027 at $1 trillion（約160兆円） valuations

   OpenAI and Anthropic filed confidential draft S-1 registration statements with the SEC on May 22 and June 1, respectively, and are expected to go public in late 2026 or early 2027. OpenAI aims for a $1 trillion（約160兆円） market cap (50 times its $20 billion（約3.2兆円） annualized revenue run rate), while Anthropic would go public at a $1 trillion（約160兆円） valuation (111 times its $9 billion（約1.4兆円） annualized revenue). Both companies plan to float only 5%-10% of their shares, which combined with market hype could cause their stocks to skyrocket on debut—even though they are already expensive relative to revenue. Investors seeking early exposure face significant risks: secondary-market shares can be voided by the companies, and exchange-traded funds tracking these firms may trade well above their actual net asset value, leaving latecomers with losses when the hype fades.

   Accredited investors (net worth over $1,000,000 excluding primary residence, or annual income over $200,000) can currently buy shares through secondary marketplaces like Forge Global, Hiive, and EquityZen, or via special-purpose vehicles and mutual funds such as ARK Venture Fund and Fundrise Innovation Fund. However, waiting for the IPO may offer better entry prices once the market prices in steep valuations and net losses.

4. 4

   Microsoft launches $2.5 billion（約4000億円） Frontier unit to embed AI engineers with enterprise customers

   Microsoft announced the creation of Microsoft Frontier Company with a planned investment of $2.5 billion（約4000億円）. The unit will place 6,000 AI engineers and specialists directly inside enterprise customer operations, working in partnership with firms such as Accenture and KPMG to accelerate large-scale AI adoption. Rather than relying solely on cloud tools, Microsoft is taking a resource-intensive, service-led approach to help large organizations move from pilot AI projects to broader deployment. This shift may strengthen long-term customer relationships if projects succeed, though the real payoff will depend on how often management connects this investment to revenue, margins, and contract wins in future updates.

   This represents a material expansion of Microsoft's AI services effort. Investors should track how the company demonstrates a return on the $2.5 billion（約4000億円） commitment through customer outcomes and business results.

5. 5

   Reddit AI licensing deals exceed $200M/year, fueling growth push

   Reddit's Q1 2026 revenue jumped 69% year-over-year, with advertising revenue growing 74% and daily active users reaching 126.8 million (up 17%). The company has signed AI licensing deals worth over $200 million（約320億円） per year combined—Google pays $60 million（約96億円） annually and OpenAI pays around $70 million（約110億円） annually—giving both access to Reddit's real-time data to train their AI models. Both deals renew in 2027. Reddit is no longer primarily a social media platform; it has become a data supplier to AI companies. As high-quality training data has become scarce, Reddit's user-generated discussions are uniquely valuable. The company is pushing for dynamic pricing rather than flat fees on future renewals, signaling that AI companies have limited alternatives and Reddit's negotiating power is rising. For investors, this licensing revenue stream, combined with strong advertising growth, offers a concrete revenue diversification beyond traditional social media.

   Reddit's average revenue per user is $5.23 globally and $9.63 in the US—already exceeding Pinterest ($7.12) and Snapchat ($9.23), and approaching Meta ($15.66). The 2027 renewal of both major AI licensing deals is a key catalyst; the outcome will signal whether the company can secure higher rates and demonstrate sustained growth in this new business line.

6. 6

   Australian AI infrastructure startup IREN lands $13.1B in mega cloud deals

   IREN, an Australian company that pivoted from Bitcoin mining to AI infrastructure, signed a 5-year $9.7 billion（約1.6兆円） AI cloud contract with Microsoft in November and a $3.4 billion（約5400億円） contract with Nvidia in May, making it one of the market's hottest "neocloud" stocks. These contracts are expected to drive IREN's revenue from $510 million（約820億円） in fiscal 2025 to $6.44 billion（約1兆円） by fiscal 2028, with adjusted EBITDA rising from $270 million（約430億円） to $4.89 billion（約7800億円）. For investors seeking faster growth than Nvidia—already a $4.7 trillion（約750兆円） company—IREN's smaller scale offers potential for outsized returns, though valuations remain compressed by near-term spending concerns.

   Analysts project IREN will achieve profitability by fiscal 2028. The company's enterprise value is $15.7 billion（約2.5兆円）, valued at seven and ten times next year's expected sales and adjusted EBITDA, respectively.