5 Major AI Stock Analyst Moves: Google Stays a Top Pick After Gemini Agent Launch
Wall Street’s latest artificial intelligence (AI) stock calls show that investors are becoming more selective about where they expect the next phase of growth to emerge. Alphabet remains a preferred AI investment for BMO Capital Markets following Google’s launch of Gemini Agent, while analysts are highlighting semiconductor companies positioned to benefit from data-centre expansion and reassessing businesses with weaker exposure to AI infrastructure.
The latest analyst moves also reveal a more cautious side of the AI trade. Citi has downgraded NXP Semiconductors, and Truist Securities has initiated coverage of IBM with a Hold rating, reflecting concerns about near-term execution, automotive demand and the timing of enterprise software spending.
For investors, the message is not that every AI-related stock will benefit equally. The next phase may increasingly depend on where companies sit in the AI supply chain, whether their revenue forecasts are achievable and how much future growth is already reflected in their share prices.
1. BMO Reaffirms Alphabet as a Top Pick After Gemini Agent Launch
BMO Capital Markets has reiterated its Outperform rating and Top Pick designation on Alphabet, Google’s parent company, after the launch of Gemini Agent for enterprise users. The product is designed to carry out multi-step workplace tasks and strengthens Google’s effort to compete in the emerging market for AI agents.
Unlike a conventional chatbot that primarily answers prompts, an AI agent is designed to help complete tasks across connected applications and workflows. This could make AI more useful to businesses that want to automate research, coding, document work and other routine processes.
BMO’s case rests partly on Alphabet’s existing enterprise relationships and its ability to combine AI models, computing infrastructure and software distribution. The brokerage highlighted connections with tools such as Microsoft Office, Salesforce and ServiceNow, alongside integrations within Google’s own product ecosystem. It also noted that around 80% of Google Cloud customers were already using Google AI products, according to the report.
Another factor is the declining cost of running AI systems. BMO said the cost per token had fallen by 98% since 2024. A token is a unit of text or other data processed by an AI model. Lower processing costs can help providers serve more customers while improving the economics of AI products, although pricing competition could also limit margins.
What investors should watch: Whether Gemini Agent translates into paid enterprise adoption, stronger Google Cloud growth and measurable revenue. A product launch can improve a company’s competitive position, but it does not automatically guarantee a meaningful financial contribution.
2. Citi Names AMD, Lam Research and Other Chip Stocks Among Its Preferred Picks
Citi has identified AMD, Texas Instruments, Lam Research, Teradyne and Synopsys as its top Buy-rated semiconductor selections ahead of third-quarter earnings. The move reflects continued confidence in AI infrastructure demand, but also a more selective approach after a strong recovery in semiconductor shares.
The bank highlighted data centres as a major source of industry demand, accounting for about 40% of the semiconductor market in its analysis. Automotive and industrial applications represented roughly another 20%, while personal-computer and smartphone demand remained comparatively weak.
The distinction matters because AI-related semiconductor demand is not distributed evenly. Advanced processors and networking chips support data-centre expansion, while semiconductor-equipment suppliers benefit when chipmakers invest in manufacturing capacity. Analog chipmakers have different demand drivers, including industrial systems, vehicles and power management.
Citi also observed that earnings-estimate upgrades had slowed in the second quarter, with further moderation expected in the third. That suggests investors may increasingly focus on which companies can deliver results above expectations rather than simply benefit from the AI theme.
The preferred names cover different parts of the industry:
AMD: Processors and accelerators used in computing and data-centre applications.
Lam Research: Equipment used in semiconductor manufacturing.
Teradyne: Automated test equipment for chips and electronics.
Texas Instruments: Analog and embedded-processing chips.
Synopsys: Electronic design software and semiconductor intellectual-property tools.
These businesses do not have identical growth prospects or risk profiles. Investors should compare order trends, customer demand, margins and valuation rather than treat the entire semiconductor sector as one trade.
3. Marvell Technology Gets Higher Price Targets as Long-Term Ambitions Expand
Marvell Technology attracted several upward price-target revisions after presenting a more ambitious long-term financial framework at its analyst day.
Management raised its fiscal 2028 revenue outlook to approximately $20 billion from $18 billion. It also outlined a fiscal 2031 revenue framework of $70 billion to $90 billion, with an $80 billion midpoint, alongside a target for earnings above $30 per share and gross margins of 56%–59%.
The company’s investment case is closely linked to connectivity: the networking and interconnect technologies that allow processors, accelerators and other components to communicate inside increasingly complex data centres.
Several analysts responded positively. TD Cowen upgraded Marvell to Buy from Hold and raised its price target to $350 from $245. Jefferies lifted its target to $450 from $325, while Stifel, B. Riley and Susquehanna also raised their respective targets. These are individual brokerage estimates, not guaranteed future share prices.
The broader significance is that AI infrastructure spending extends beyond the chips that perform calculations. Faster networks and high-bandwidth connections become increasingly important as computing workloads expand.
However, Marvell’s long-term framework extends several years into the future. The company will need to convert its opportunities and project pipeline into actual sales, earnings and cash flow. Execution delays, customer concentration, competition and changes in data-centre spending could all affect the outcome.
What investors should watch: Quarterly revenue growth, networking demand, gross margins and evidence that the company is progressing towards its longer-term targets.
4. Citi Downgrades NXP Semiconductors as Automotive Demand Remains Weak
Citi downgraded NXP Semiconductors to Neutral from Buy and cut its price target to $260 from $370. The brokerage expects weaker earnings-estimate revisions than those of some semiconductor peers as the industry moves further into 2027.
NXP has substantial exposure to automotive electronics, including chips used in vehicle systems. That exposure can be a strength when vehicle production, electronics content and demand are growing. It can become a disadvantage when automakers and suppliers face weaker demand or delay purchases.
According to Citi’s analysis, automotive accounted for 56% of NXP’s revenue, while data centres represented only around 3%–4%. This leaves the company less directly exposed to the infrastructure spending that is currently supporting some other chipmakers.
Citi also reduced its earnings-per-share estimates for 2027 and 2028, citing weaker assumptions for automotive and industrial markets. Its downgrade illustrates an important point: being a semiconductor company is not enough to make a stock a direct beneficiary of the AI boom.
NXP could still benefit from longer-term growth in vehicle electronics and industrial applications. But those structural opportunities must be weighed against near-term demand conditions and valuation.
Investors should monitor automotive production, customer inventory levels, industrial demand and changes to earnings forecasts before drawing conclusions about a recovery.
5. Truist Starts IBM Coverage at Hold, Citing Execution Risks
Truist Securities initiated coverage of IBM with a Hold rating and a $240 price target. The brokerage acknowledged IBM’s position in enterprise computing but raised concerns about how quickly spending on AI infrastructure will translate into stronger demand for the company’s software and services.
IBM operates across several areas relevant to enterprise AI, including mainframe computing, hybrid-cloud software, Red Hat and consulting. These businesses could help organisations integrate AI into existing systems rather than replace their entire technology infrastructure.
The challenge is timing. Companies may be directing budgets towards servers, storage and memory before increasing spending on software projects. If customers postpone software decisions, IBM could struggle to convert its established relationships into the growth investors expect.
The report noted that software represented 45% of IBM’s revenue and 61% of segment pre-tax income, with recurring revenue providing an important foundation. It also reported that IBM had lowered its full-year constant-currency revenue growth guidance to 4%–5% from more than 5%.
Truist identified several potential signs of improvement: delayed contracts closing, stronger transaction-processing growth and better organic software expansion.
For investors, the distinction is between having a credible role in enterprise AI and demonstrating that the opportunity is producing stronger financial results. IBM’s established customer base may be an advantage, but execution and spending cycles remain important risks.
What These Analyst Moves Mean for AI Investors
The five developments point to a more differentiated AI market. Alphabet is being favoured for its ability to combine AI models, cloud infrastructure and enterprise distribution. Marvell and selected semiconductor companies are drawing attention because AI expansion requires more than processors alone. NXP’s downgrade and IBM’s Hold rating, meanwhile, highlight the importance of near-term demand and execution.
Three themes stand out:
AI infrastructure remains a key investment focus. Data centres need processors, networking equipment, semiconductor manufacturing tools and supporting software.
Revenue conversion matters more than announcements. Product launches and ambitious long-term targets must eventually translate into sales, margins and cash flow.
Valuation and business exposure still matter. Two companies can both be associated with AI yet face very different growth prospects, risks and earnings cycles.
For Indian investors following global technology stocks, these developments offer a useful way to assess the AI sector without assuming that every company linked to artificial intelligence will benefit equally. Brokerage ratings are opinions based on particular assumptions, and price targets can change as new financial results emerge.
The next major test will be whether upcoming earnings reports confirm the demand assumptions behind analysts’ forecasts. Alphabet’s enterprise AI adoption, semiconductor order trends, Marvell’s execution against its long-term targets and NXP’s automotive recovery will be important indicators to monitor.
Follow the blog for more updates on AI stocks, semiconductor companies, global markets and business trends.
This article is for informational and educational purposes only and should not be considered investment advice.