Street Calls of the Week

Investing.com -- Here are the biggest analyst moves in the area of artificial intelligence (AI) for this week.

InvestingPro subscribers always get first dibs on market-moving AI analyst comments. Upgrade today!

Get deeper analyst insights and top AI-driven stock ideas by upgrading to InvestingPro - get 55% off today

Rothschild turns cautious Microsoft, Amazon as GenAI returns fall short

Earlier this week, Brokerage firm Rothschild is turning more cautious on the hyperscalers, downgrading Microsoft and Amazon to Neutral as it argues that the financial profile of GenAI no longer supports premium valuations.

"It is time to take a more cautious stance on the hyperscalers and move beyond the industry’s reassuring ‘trust us – GenAI is just like early cloud 1.0’ narrative, which looks increasingly misplaced," analyst Alex Haissl said in a note.

His core argument is that the GenAI build-out demands vastly more capital while delivering meaningfully lower returns.

Haissl says “GPU deployments require roughly six times more capital to generate the same cloud 1.0 value, with risks skewed to the downside,” and estimates that each dollar invested in GenAI infrastructure yields only about twenty cents of net present value, versus roughly $1.4 for mature cloud 1.0 projects.

For Microsoft, he sees the quality of growth deteriorating as Azure becomes more exposed to lower-value workloads. The once-powerful AI shield is fading, he argues, because GenAI economics flow toward model developers rather than platform owners.

In his view, “value increasingly flows to model providers like Anthropic or OpenAI,” weakening the leverage Microsoft previously enjoyed. The analyst also points to the risk of value leakage inside Office 365, where AI features embed more cost than monetization.

Amazon faces many of the same pressures. Although AWS benefits from custom silicon and deeper vertical integration, Haissl characterizes GenAI as the sector’s “low value growth engine,” accounting for more than half of cloud revenue expansion but requiring sustained and heavy capex.

The constraint, he writes, is “the capex required relative to the growth delivered,” with spending levels that would not exist without the AI wave.

But despite the downgrade, Haissl stresses that this is not a bearish call. Growth should remain healthy over the coming year as new capacity ramps, but he argues that upside is already priced in and that GenAI no longer offers a path back to the attractive economics of cloud 1.0.

“We no longer see a credible path back to cloud 1.0 economics,” the note concludes, a disconnect he sees at the heart of his more cautious stance.

Rothschild lowered its Microsoft target to $500 from $560 and kept its Amazon target at $250.

Google could regain AI momentum as Gemini 3 launch nears: Mizuho

Google’s competitive position in AI may be due for a boost, according to Mizuho, which argues that the market’s fixation on OpenAI’s threat has overshadowed signs of improving fundamentals for Alphabet.

Analysts led by Lloyd Walmsley say sentiment could turn more constructive as Google prepares to introduce its next major model.

They highlight that the “imminent roll-out of Gemini 3 could further tilt the sentiment for Alphabet shares towards AI-winner, at least near term,” supported by steadier search activity and rising engagement with Google’s own AI features.

Mizuho’s tracking shows ChatGPT’s usage momentum has cooled since spring, with slower time-spent trends on desktop and mobile, softer app engagement and another expected monthly decline in global downloads.

In contrast, Gemini appears to be gaining traction inside Google’s ecosystem. Browser traffic and token-processing data suggest widening usage, while the combined market share of GPT and Gemini has climbed from 7% to 13% since late 2024, with recent gains “skewing toward Gemini” as Google tools see stronger uptake.

"The trend indicates broad usage of Google AI products across Search, Gemini, and other channels, which has grown, and could continue growing exponentially driven by Google’s product ecosystem," the analysts wrote in a note.

Google’s AI Mode in Search has helped return search traffic growth to positive territory, supporting early signs that Alphabet’s strategy is stabilizing query share. Gemini’s browser share is rising as well, helped by what analysts describe as the product “likely getting the most traffic “through AI Mode in core search.”

Gemini 3 is expected to bring advances in multimodality, reasoning, and automation, including real-time video understanding, deeper context windows, a new default reasoning mode, tighter Workspace integration and an “Agent Mode” designed for browser-level task execution. Developer signals also point to on-device AI and closer integration with Veo 3.1.

Mizuho believes these upgrades could help narrow the gap with OpenAI and potentially shift the competitive balance in enterprise automation. While the long-term race remains uncertain, the broker sees room for overhangs to fade as Alphabet rolls out its next AI wave.

“In the near term, our bias is to think that Google continues to benefit on the core business fundamentals from AI and rolls out impressive AI product of its own - most imminently Gemini 3 - that continues to swing the pendulum more towards AI winner and less towards AI loser,” the analysts said.

Oppenheimer initiates IBM at Outperform on software-driven growth, AI upside

Oppenheimer this week began coverage of IBM with an Outperform rating, arguing that the company’s pivot toward software and AI is setting up a period of sustained revenue and margin gains.

The broker set a 12- to 18-month price target of $360 on the stock.

Analyst Param Singh says the constructive stance reflects confidence in IBM’s software trajectory, with the portfolio expected to deliver lasting double-digit revenue growth.

He believes this momentum will be “driven by strength in Automation (primarily HashiCorp) and improving growth in RedHat,” two pillars of IBM’s push toward a higher-margin, software-centric model.

Consulting should grow at a steady low single-digit pace as application development and management demand improves.

Oppenheimer also sees additional upside in AI. Singh points to “additional revenue optionality with creation and management of AI applications (incl. Generative AI),” arguing that IBM is well placed as enterprises scale their use of AI tools.

This should support “strong expansion activity with existing customers,” with software mix gains feeding continued improvement in gross margins.

Pre-tax margins are also expected to rise as IBM benefits from its product mix and operational leverage. Singh writes that the shares could earn a higher valuation as investors gain more clarity on the company’s transformation.

“The stock should also re-rate higher when IBM’s pivot to software is more widely appreciated," Singh wrote.

Raymond James sees ’secular boom’ in semiconductors as AI reshapes the sector

Raymond James has resumed coverage of seven large semiconductor names, arguing that generative AI has “transformed a typically cyclical market sector of semiconductors into a secular boom.”

Analyst Simon Leopold wrote that enterprises are still in the “early stages of adoption,” while AI-factory builders plan to scale their investments and construct ever-larger facilities.

He says its full-stack data-infrastructure work points to rapid innovation across compute, networking, storage, optics and infrastructure software.

Leopold also highlights power-supply constraints and stresses that technologies such as “co-packaged optics” and “CoWoS / Chip on Wafer on Substrate” will be essential to expanding high-performance clusters.

Coverage was resumed on NVIDIA (Strong Buy), Marvell (Strong Buy), Broadcom (Outperform), AMD (Outperform), ARM (Market Perform), Astera Labs (Market Perform) and Intel (Market Perform).

Leopold describes semiconductors as “foundational elements” of “an unprecedented structural shift in the technology landscape.”

The analyst says NVIDIA “leads in AI and is a core holding,” citing its software stack, developer ecosystem and “full-stack systems approach.” He expects “continued upward estimate revisions” and models peak Blackwell orders of 7.8 million units in fiscal 2027.

Marvell “faces more skepticism,” but is viewed as a “share gainer” in custom ASICs and optical DSPs, with management aiming for about 20% data-centre ASIC share by 2028.

Broadcom, meanwhile, is described as “a share gainer in the AI complex,” and AMD as “best positioned to compete with NVIDIA” in merchant GPUs. The note also flags ARM’s traction in data centres and Astera Labs’ “first-mover advantage” in high-speed interconnects.

For Intel, Leopold argues execution must “catch up,” and that its loss-making foundry unit “remains an albatross.”

BofA turns more bullish on Sandisk, lifts target to Street-high $300

Bank of America (BofA) has grown more bullish on Sandisk after meeting with chief executive David Goeckeler and chief financial officer Luis Visoso, raising its price target to $300 — the highest on the Street — while reiterating a Buy rating.

The bank said it “walked away more bullish on the stock” following discussions in San Jose.

BofA says management expects the NAND market to stay “undersupplied through at least end" of 2026, a setup it believes should support pricing as demand strengthens.

The bank points to six key growth drivers, including “strong demand growth driven by datacenter and AI,” low industry inventories, Sandisk’s expanding eSSD output and the potential for share gains as the company accelerates its BiCS8 transition. It also highlighted Sandisk’s “increasing visibility from customers.”

Despite the tight supply backdrop, management has “no plans to ramp capacity,” according to BofA, noting that customers are now “eager to lock in volumes.”

Sandisk is negotiating longer-term agreements balancing volume commitments and pricing, drawing a parallel to Goeckeler’s prior discipline efforts at Western Digital. At the same time, the company is watching closely for signs of over-ordering.

Management described market conditions as “tepid” even with mid-teens demand growth and said first-quarter trends are likely to remain seasonally soft.

The company plans to maintain a net cash position, with Visoso preferring about “$1.3bn of cash on hand” for operations. BofA expects shareholder returns to “focus on buybacks.”

BofA raised its valuation multiple to 2.7x 2027 estimated price-to-book, arguing that Sandisk merits parity with peers as its eSSD ramp and BiCS8 migration “are already driving profitability higher.”

Investing.com -- OpenAI, the San Francisco–based artificial intelligence developer behind ChatGPT, is preparing a broader push into the enterprise market as its business customer base and revenue mix shift sharply toward corporate use. 

CEO Sam Altman recently said the company is “quite excited about our upcoming enterprise offering,” a message that coincides with stronger enterprise monetization efforts and growing investor attention to how OpenAI plans to scale beyond consumer subscriptions.

UBS said that OpenAI’s enterprise revenue has grown ninefold year over year, moving from roughly 30% of total revenue at the beginning of 2025 toward a 60/40 split with the consumer business. 

The firm now has more than one million enterprise customers, according to the company’s CFO. She also told the Wall Street Journal that ChatGPT for Enterprise has surpassed a $1 billion run rate, with margins resembling traditional enterprise software. 

UBS notes that ChatGPT for Business seats have climbed to more than 7 million, up from 5 million in August and 3 million in June, a 40% gain in two months.

Coding represents one of the clearest areas where OpenAI is likely to move next, UBS says. 

Altman identified coding as the company’s next major revenue driver, and the analysts highlighted the launch of the Codex assistant and the GPT-5-powered Aardvark tool, which searches for security vulnerabilities in software. 

The brokerage says these efforts position the company more directly against developer-focused incumbents such as Atlassian, GitLab and Microsoft’s GitHub Copilot. 

UBS notes that competitors Anthropic and Cursor are already near or above $1 billion in annualized revenue from coding tools, intensifying competitive pressure in the market.

The analysts also point to employee productivity software as a likely expansion area. With 7 million business users, OpenAI now has a meaningful foothold in a market dominated by Microsoft Office 365 and Google Workspace, which together generate $70–75 billion in annualized revenue. 

UBS says extending ChatGPT beyond search into workflow tasks could allow OpenAI to compete more directly in productivity use cases, particularly among Google’s small-business and education segments. 

The brokerage notes OpenAI’s existing integrations with Google Workspace and highlights Anthropic’s introduction of a Claude-powered Excel product as evidence of growing competition in this category.

OpenAI is also positioned to deepen its “scaffolding” strategy by offering tools that help enterprises build their own AI applications, UBS said. The analysts say the company’s AgentKit, introduced at DevDay, moves it closer to hyperscaler platforms such as Azure AI Foundry, AWS Bedrock and Google Vertex AI. 

UBS adds that reinforcement fine-tuning and model-customization services may become more prominent as enterprises integrate proprietary data. The acquisition of Statsig could also support analytics and experimentation capabilities within OpenAI’s ecosystem.

UBS flags industry verticals as another emerging focus. The CFO said OpenAI is deepening collaborations in science, healthcare, tech and consumer goods, with deals tied to high-value research breakthroughs. 

UBS also cites the formation of “OpenAI for Science” and new hires in healthcare leadership roles as signals of growing vertical strategy, noting that Anthropic has taken similar steps with financial-services-focused models.

Despite the broad opportunity, UBS says compute constraints may influence timing. The brokerage cites OpenAI’s applications chief, who told Wired that the “real question becomes, will we have the compute to deliver that?,” a limitation UBS identifies as potentially shaping how fast the company can scale new enterprise offerings.

Investing.com -- Block’s latest analyst day marks a “positive inflection,” according to Mizuho, which argues the company may now be better positioned to push Bitcoin toward mainstream, everyday use. 

Analyst Dan Dolev wrote that the event was notable not just for new AI products such as ManagerBot and MoneyBot, but also for the company’s candid acknowledgment of past missteps. 

According to Dolev, Block’s CEO opened with “mea culpas ranging from lack of adequate product velocity to prior organizational challenges.”

But the firm’s optimism centres on Block’s long-term Bitcoin strategy. Mizuho said it is “even more thrilled about Block’s opportunity to lower the cost of acceptance by utilizing Bitcoin rails,” highlighting faster banking penetration, the “surprisingly strong health and high returns” of the company’s lending operations, and what it called a “broader infrastructure opportunity embedded in a one-stop-shop for banking + wallet + POS + Bitcoin + mining.”

The analyst said the economic incentives for Bitcoin payments “seem real.” 

“The Bitcoin payment rails, that XYZ is piloting, could potentially be transformational,” stated Dolev. “Cash Apps’s 58 million users are now able to use the Lightning Network for payments. We believe that this is potentially one of the most consequential changes in payments over the last 50 years.”

He explained that Lightning offers near-instant, low-cost payments and eliminates chargebacks and intermediaries, which “could encourage merchants to offer discounts for BTC payments, passing savings to customers.”

“We believe this opportunity is incremental and transformative and Block is at the forefront of this shift today,” Mizuho argues.

Furthermore, Dolev said that “building viral networks is key advantage & focus,” and that for Block, “all roads lead to primary banking” as it integrates financial services around Cash App and its merchant ecosystem.

Mizuho maintained its Outperform rating, lifting its price target for the stock to $100 from $88 per share, concluding that it is “incrementally upbeat about XYZ” as Block pursues what it sees as a viable path to making Bitcoin a routine means of payment.