AI Daily: AI model start-up Reflection AI raises $2B

Catch up on the top artificial intelligence news and commentary by Wall Street analysts on publicly traded companies in the space with this daily recap compiled by The Fly:

NEW FUNDING ROUND: AI Reflection AI, a year-old American start-up, is betting that it can create a Western equivalent of DeepSeek. And it has raised billions to do so, The New York Times’ Michael J. de la Merced reports. Reflection AI plans to announce on Thursday that it has raised $2B in new funding, in a round led by Nvidia (NVDA), the chipmaker at the heart of the AI boom, the author notes. The deal values Reflection AI at $8B, including the new money, significantly more than the $545M the company was valued at in March, according to the data provider PitchBook.

LEADERSHIP POSITION: Morgan Stanley reports that the firm’s Q3 CIO survey found that artificial intelligence and machine learning remain at the top of the CIO priority list. Given hyperscalers remain the favored vendor group for CIOs in supporting deployment efforts for GenAI and LLMs, as well as Microsoft’s (MSFT) deep integrations across the software ecosystem, vast scope of products and significant and increasing AI infrastructure investments, the firm argues that Microsoft “remains in the leadership position” to capture incremental share of GenAI spend and IT budgets as workloads move to the cloud. The firm keeps an Overweight rating and $625 price target on Microsoft shares

CLOUD PROVIDER: Google (GOOGL) will serve as the official cloud provider of the Los Angeles 2028 Olympic and Paralympic Games and has joined LA28 as a founding partner in a deal that also extends to Team USA and NBCUniversal’s (CMCSA) U.S. coverage, Reuters’ Rory Carroll reports. The partnership will use Google’s cloud infrastructure and AI tools, including its Gemini model and new Google Search features such as “AI Mode” to help fans discover information during NBCUniversal’s multi-platform broadcasts and to support LA28 operations.

PRICE TARGET RAISED: Cantor Fitzgerald raised the firm’s price target on Nvidia to $300 from $240 and keeps an Overweight rating on the shares after hosting meetings with management. The multi-trillion artificial intelligence infrastructure buildout remains in the early inning with just the hyperscalers “providing significant line-of-sight into hundreds of billions of demand for the next handful of years,” the analyst tells investors in a research note. The firm this exclude other drivers such as neo-clouds, enterprise, and physical AI. “This is not a bubble,” proclaims Cantor, which thinks the investment cycle benefiting Nvidia is in the early stages. The firm says its confidence in both the growth in AI infrastructure and Nvidia’s share into the 2030 has “increased considerably” post the meetings. Cantor continues to see $8 in earnings per share for Nvidia as “very doable” into 2026, followed by $11 in 2027, well above the consensus of $7.36. The stock remains the firm’s top pick.

AI INFRASTRUCTURE EXPOSURE: Baird initiated coverage of Oracle (ORCL) with an Outperform rating and $365 price target. The firm views Oracle as the “AI juggernaut for the information age.” The company is well positioned to benefit from the accelerating spend on artificial intelligence infrastructure as well as convergence of AI, data and use-cases that will emerge from the move from training to inference, the firm tells investors in a research note. This “virtuous circle” can support a premium valuation for Oracle and move the shares higher, contends Baird.