AI Stocks & Markets

Today's Stories

1. 1

   DeepSeek's new reasoning model reaches #2 ranking among open-weights models, signaling intense competition in AI development.

   DeepSeek released an upgraded reasoning model that has become the #2 open-weights reasoning model. The model operates with 27% of FLOPs compared with DeepSeek-V3.2 and uses 83.9 GiB of memory. Open-weights models are increasingly competitive alternatives to proprietary AI systems, and DeepSeek's rapid advancement demonstrates how quickly the landscape is shifting. This reflects broader industry momentum toward making powerful reasoning capabilities accessible outside closed ecosystems.

   The model's efficiency gains—operating on significantly fewer computational resources than its predecessor while achieving higher rankings—may influence how other developers approach reasoning model design going forward.

2. 2

   AppLovin offers a better risk-reward profile than ARM Holdings for AI investors, combining stronger earnings growth with a more reasonable valuation.

   AppLovin is expected to deliver 42% year-over-year sales growth and 58% earnings-per-share growth for 2026, driven by its AI-powered advertising platform and expansion into web and e-commerce advertising. ARM, by contrast, is projected to grow sales 21% and earnings per share 19%, supported by data-center royalty revenue that more than doubled year over year and close to 50% share among leading hyperscaler cloud compute deployments. AppLovin trades at a forward 12-month P/E of 25.64X, below its median of 34.89X, suggesting a more attractive valuation. ARM trades at 175.65X, well above its median of 123.38X, reflecting steep investor expectations. AppLovin's stronger earnings momentum and operational efficiency in converting growth to profit make it the more compelling near-term opportunity despite ARM's quality ecosystem and exposure to long-term AI and cloud infrastructure trends.

   AppLovin's self-serve Axon Ads rollout is expected during the first half of 2026, which the company sees as a potential turning point for broadening adoption among non-gaming advertisers. The expansion of its generative creative technologies, including interactive landing-page generation and future video-ad tools, is expected to strengthen campaign performance and improve conversion metrics.

3. 3

   AI industry factions clash over regulation in New York House race, as political spending reveals a deepening rift between pro-regulation and anti-regulation camps.

   In a crowded Democratic primary for a New York congressional seat, two opposing factions within the AI industry have spent heavily on ads. A political group underwritten by investors in OpenAI spent more than $7 million（約11億円） on ads against New York Assemblyman Alex Bores, who authored a sweeping state-level AI regulation bill. Political groups partly funded by Anthropic, which makes the chatbot Claude, responded by spending more than $10 million（約16億円） to boost Bores' candidacy. Bores left Palantir citing ethical concerns and has made AI regulation central to his campaign platform. The industry's public split—with OpenAI-backed groups opposing him and Anthropic-backed groups supporting him—suggests a genuine fault line within the AI sector over how Congress should approach regulation. The outcome will offer a measure of the political influence each faction can wield.

   The primary results on Tuesday will show whether the pro-regulation side can overcome OpenAI-backed spending in a competitive race. The outcome may signal which regulatory posture has stronger backing among Democratic voters and industry leaders as AI policy heads to the federal level.

4. 4

   Tech stocks plunged for a second day as investors demand proof that AI spending will actually deliver profits, not just promises.

   The Nasdaq Composite fell 2.2% on Tuesday, marking a second straight day of losses. Semiconductor and cloud companies including Nvidia (down 4.2%), Broadcom (down 3.1%), Meta Platforms, and Microsoft have all entered bear market territory—a 20% drop from their recent peaks. SpaceX shares fell after the company's June IPO sent them above $200; they have since retreated amid concerns about whether the company can justify its valuation exceeding $2 trillion（約320兆円）. For months, Wall Street bet heavily that AI investments would drive faster revenue growth and higher profits. But new data from the Bank of America Institute shows only about 3% of its customers—mostly those with more than $125,000 in annual income—pay for AI services, spending a median of $20 per month. Investors are now demanding concrete evidence that the enormous spending on AI will translate into actual returns, rather than accepting the technology's potential on faith alone.

   The number of households paying for AI services jumped 38% since 2024, and Bank of America Global Research expects the U.S. market could scale to $75 billion（約12兆円） annually as AI becomes embedded across productivity, search, entertainment, and personal assistant use cases. At the same time, anxiety over potential Federal Reserve rate hikes later in 2026 is also weighing on the market, with traders betting on a nearly 90% chance of at least one rate increase by year-end.

5. 5

   Seagate stock has surged 297.98% year to date on AI-driven storage demand, but analysts say the easy gains are likely behind it at current valuations.

   Seagate stock rose from $274.90 on December 31, 2025 to $1,094.04 on June 22, 2026, driven by strong Q3 earnings—adjusted EPS of $4.10 versus $3.50 expected, revenue up 44.07% year over year, and management raising its annual revenue growth target to a minimum of 20% over the next few years. The company reported exabyte-scale supply agreements (large-scale data storage contracts) with major cloud and hyperscale customers, with nearline capacity almost fully allocated through calendar 2027. Seagate's rally reflects investor confidence in hard drive demand from AI data centers, a sector many thought obsolete. However, the stock now trades at a P/E of 102 and roughly 88x forward earnings, well above the $898.09 analyst consensus target. Company insiders, including CEO Dave Mosley, sold shares in mid-June at $880.19, though these were pre-planned transactions. This combination of stretched valuation and insider sales suggests the structural AI thesis remains sound, but near-term upside may be limited.

   The 24/7 Wall St. price target is $1,010.11 over the next 12 months, implying roughly 7.7% downside from current levels, with a hold recommendation at 90% confidence. Key risks include whether Seagate can qualify its HAMR (heat-assisted magnetic recording) technology with remaining hyperscalers on schedule and whether Q4 results show softening in pricing or miss the mid-20% growth target for exabyte shipments. A pullback to the $850–$900 range would look more attractive on a risk/reward basis if those milestones hold.

6. 6

   Google executive Demis Hassabis says the company is retaining top AI talent despite departures, citing its unmatched computing infrastructure and unified research organization as competitive advantages.

   DeepMind chief Hassabis expressed confidence that Google is still winning the competition for AI researchers, even as prominent scientists like Noam Shazeer and John Jumper have left to start new ventures. The company merged Google Brain and DeepMind in 2023 to create a centralized research organization under Hassabis's leadership focused on advancing the Gemini ecosystem. In AI research, individual breakthroughs can unlock billions in revenue, making top researchers highly sought after. Venture investors view researcher departures as a sign of potential 'brain drain' and use human capital as a leading indicator of which company will dominate AI development. Hassabis's argument rests on Google's structural advantages—its vast data ecosystem, custom hardware (Tensor Processing Units), and computing power—which he believes remain unmatched recruiting tools for frontier model development.

   Google's ability to retain and attract elite researchers will shape whether it can compete with nimble startups that offer massive equity upside. The centralized research structure created by the 2023 merger positions the company to pursue both frontier AI model development and fundamental scientific discovery like AlphaFold.