Investor attention is focused on Alphabet's second-quarter earnings report, with markets looking for evidence that the company's aggressive artificial intelligence (AI) investments are translating into sustainable revenue growth.

Alphabet is scheduled to report its second-quarter results after US markets close on July 22 (around 5:00 a.m. Korea time on July 23). While the company's strong first-quarter performance reassured investors, the second quarter has raised fresh concerns over delays to the flagship Gemini 3.5 Pro AI model, the departure of key AI researchers and rising infrastructure investment costs.

Alphabet will webcast its earnings conference call at 5:30 a.m. Korea time on July 23 through its investor relations website and YouTube channel. According to Reuters, investors are primarily looking to determine whether Alphabet's record data center investments and Gemini AI strategy are generating tangible financial returns. As a result, growth in the company's cloud business has become the most closely watched metric.

Cloud revenue expected to grow 64% year over year

According to estimates from LSEG, Alphabet is expected to post revenue of $116.93 billion for the April-June period, up 21.3% from a year earlier. Google Cloud revenue is forecast to increase 64%, while advertising revenue is expected to grow 13.7%.

Other market consensus estimates are broadly similar. FactSet expects revenue of about $117 billion, with Google Cloud revenue growing 63% to 70%. Visible Alpha forecasts revenue of $117.19 billion, representing growth of more than 20%, while projecting Google Cloud revenue of approximately $22.5 billion, up more than 65%, and advertising revenue from Search and related businesses of roughly $63.29 billion.

Advertising, including Google Search and YouTube, remains Alphabet's largest source of revenue, generating more than three times the sales of its cloud business. Nevertheless, investors are paying closer attention to cloud growth because Alphabet's enormous capital expenditures (CAPEX) must ultimately be justified by accelerating revenue. The company has guided for $180 billion to $190 billion in CAPEX this year.

Even if consensus expectations are met, representing growth from both a year earlier and the previous quarter, investors are likely to demand more. Market participants expect that even a solid earnings report could disappoint if Google Cloud's growth rate fails to exceed expectations.

Cloud performance seen as a barometer of the global AI investment cycle

The results are also expected to influence sentiment across the broader AI ecosystem. If Google Cloud growth surpasses expectations, investors may interpret that as confirmation that the AI infrastructure investment cycle remains intact, supporting continued demand for graphics processing units (GPUs) and benefiting suppliers of optical modules and networking equipment.

The key question is whether Alphabet has maintained investor confidence in its AI strategy. During the earnings call, analysts are expected to ask whether management will maintain its full-year CAPEX guidance. Investors are also seeking updates on the launch schedule for Gemini 3.5 Pro, Google Cloud's order backlog and expansion of the company's in-house AI processors, or tensor processing units (TPUs).

A significant portion of Alphabet's massive CAPEX is being directed toward AI model development. On July 16, reports that Google had delayed the launch of a new version of Gemini, the centerpiece of its AI strategy, sent Alphabet shares down nearly 4%.

Investor sentiment failed to recover afterward, with the stock falling another 2.2% in the following trading session on July 17. Some analysts estimated that Alphabet's market capitalization declined by roughly $200 billion on July 16 alone.

The decline reflected not only concerns over the Gemini delay but also broader investor anxiety about intensifying AI competition and the scale of Alphabet's AI-related capital spending.

At Google I/O 2026 in May, Google introduced the lightweight Gemini 3.5 Flash model and said Gemini 3.5 Pro, its flagship model, would launch in June. The product, however, has yet to be released, with its launch repeatedly delayed.

While Google has rolled out lighter and lower-cost models, including Gemini 3.6 Flash, Gemini 3.5 Flash-Lite and Gemini 3.5 Flash Cyber, the flagship Gemini 3.5 Pro remains unavailable. Bloomberg and other media outlets have reported that the delay could extend for several months.

Gemini flagship delayed over coding performance

The primary reason for the delay is believed to be coding performance. According to Bloomberg and Reuters, internal testing showed that Gemini 3.5 Pro failed to meet Google's expectations in coding tasks, prompting the company to postpone its release. Coding capability has become a defining benchmark for frontier AI models because it is closely linked to agentic AI functionality.

Google is therefore believed to be delaying the launch until the model reaches its internal performance targets.

Adding to investor concerns has been the recent departure of several key AI researchers.

While movement of AI talent between major technology companies and startups is common, one of the most significant departures was Noam Shazeer, one of the co-leaders of the Gemini project. Shazeer joined OpenAI, prompting Reuters to describe the move as the loss of a key leader behind Gemini.

Another high-profile departure was John Jumper, chief scientist at Google DeepMind, who joined Anthropic. Jumper shared the 2024 Nobel Prize in Chemistry with Google DeepMind CEO Demis Hassabis for developing the AI protein-structure prediction system AlphaFold.

In addition, two core Gemini researchers, Jonas Adler and Alexander Pritzel, joined Anthropic in June. Combined with the delay of Gemini 3.5 Pro, these departures have intensified investor scrutiny of Google's competitiveness in AI models at a time when rivals continue introducing increasingly capable systems.

Following Anthropic's Claude Fable 5, competitors including OpenAI with its GPT-5.6 family, xAI with Grok 4.5 and Meta with Muse Spark 1.1 have released US models with significantly enhanced coding and agentic capabilities, while China's Moonshot AI unveiled Kimi K3 on July 16, sending shockwaves through the US AI industry. Gemini 3.5 Pro will need to demonstrate performance at least on par with those offerings.

AI search impact on advertising also under scrutiny

Another major topic expected to dominate the earnings call is the outlook for Alphabet's advertising business.

Advertising remains the company's primary revenue source. Revenue from Search, YouTube and related businesses is expected to approach $70 billion this quarter, although growth is projected at only 13% to 14%. Investors are increasingly focused on whether AI-powered search features, including AI Overviews and AI search advertising, are cannibalizing Google's traditional advertising business or creating new monetization opportunities.

One potential catalyst for improving investor sentiment is Google's next-generation AI chip strategy. According to The Information, Alphabet is developing a new AI server chip under the codename Frozen v2.

The chip is reportedly not a direct successor to Google's existing TPUs but a specialized processor designed to complement them by improving inference performance for Gemini models.

According to The Information, Frozen v2 is intended to implement certain Gemini functions directly in silicon, allowing AI models to run more efficiently on hardware. Reuters reported that Google declined to comment on the report.

Wall Street analysts continue to view Alphabet's greatest competitive advantage as its full-stack AI ecosystem and consumer base. The company uniquely combines Search, YouTube and Android with a vast global consumer base, proprietary AI models, Google Cloud and custom AI chips under one roof. Analysts argue that no other AI or cloud provider—including Amazon Web Services (AWS), Microsoft, Meta, OpenAI or Anthropic—currently possesses such a comprehensive and integrated AI ecosystem.

Alphabet Inc. reported second-quarter revenue of $119.8 billion, slightly ahead of market expectations and up 24% from a year earlier. Revenue from its closely watched cloud business reached $24.8 billion, posting 82% growth and significantly exceeding analysts' consensus estimates. Despite the strong results, Alphabet shares fell about 4% during trading after the company announced plans to further increase capital expenditures (CAPEX).

​During its second-quarter earnings conference call, which began at 5:30 a.m. on July 23, Alphabet raised its full-year CAPEX outlook to $195 billion-$205 billion, up by $15 billion from its previous guidance of $180 billion-$190 billion.

​Alphabet shares initially fluctuated following the earnings release and traded little changed before declining nearly 5% in after-hours trading after the increased investment plan was announced. Reuters reported that investor concerns were driven not only by the higher CAPEX outlook but also by continued delays to the company's flagship artificial intelligence (AI) models.

​Chief Financial Officer Anat Ashkenazi defended the increase in capital spending, saying demand continues to exceed the company's investment pace. She said aggressive investment in AI data centers enabled Alphabet to secure computing capacity faster than expected, helping Google Cloud achieve its strongest quarterly growth on record.

​Alphabet's second-quarter revenue of $119.8 billion exceeded the LSEG consensus estimate of $116.9 billion and the FactSet and Visible Alpha forecasts of $117.0 billion-$117.1 billion by about 2.3%. Total advertising revenue came in at $81.6 billion, slightly above the LSEG estimate of $81.1 billion.

Google Cloud significantly outperformed analysts' expectations, with revenue rising 82% year over year to $24.77 billion, representing a 10.3% upside surprise. The advertising business remained solid and largely met consensus expectations.

​However, Alphabet's Free Cash Flow (FCF) turned negative for the first time. FCF is calculated by subtracting CAPEX from Operating Cash Flow (OCF), which represents cash generated from the company's core business operations.

​Alphabet consumed $5.9 billion in cash during the quarter. Reuters cited Thomas Monteiro, Senior Analyst at Investing.com, as saying that increasing capital expenditures after posting the company's first negative FCF was inappropriate.

​During the conference call, Alphabet also highlighted the commercialization of its Tensor Processing Unit (TPU). According to Ashkenazi, the company generated its first TPU sales revenue during the quarter, although meaningful revenue from commercial contracts is expected to begin next year.

​Business Insider reported on Alphabet's second-quarter earnings. Alphabet Chief Executive Officer Sundar Pichai said the Gemini app now has 950 million monthly active users, approaching the scale of OpenAI's ChatGPT.

Reuters reported that while Google Cloud has positioned Alphabet as one of the primary beneficiaries of the AI boom, investors remain concerned about delays in the rollout of Gemini's latest AI models and increasing competition from rival systems. Analysts also directed numerous questions during the conference call regarding Alphabet's AI strategy.

​Pichai acknowledged, as previously reported by international media, that the company still needs to improve coding performance.

​"We continue to lead in many frontier areas," Pichai said. "There are also areas where we believe further improvement is needed, particularly coding and agentic coding."

​He added that Alphabet continues testing Gemini 3.5 Pro while training Gemini 4, saying, "We are dedicating significant computing resources and effort to Gemini to maintain our competitiveness."

​Pichai also said the Gemini app has reached 950 million monthly active users, while daily active users have tripled over the past year.

​Meanwhile, Microsoft Corp. and Amazon.com Inc. are scheduled to report quarterly earnings next week.

CocktailIO, an artificial intelligence (AI) and cloud computing company, has been selected as a consulting provider for the advanced consulting track of the South Korean government's 2026 Cloud Service Adoption and Expansion Program for Small and Medium-sized Enterprises, led by the Ministry of Science and ICT and the National IT Industry Promotion Agency (NIPA).

The company said it has already completed its first consulting match following its designation as a program provider, demonstrating the competitiveness of its cloud consulting capabilities at an early stage.

Building on its experience in infrastructure optimization, CocktailIO plans to offer end-to-end cloud services spanning consulting, solution deployment and operational management. The company aims to help participating businesses implement cloud migration strategies tailored to their individual operating environments.

Under the program, CocktailIO will provide advanced consulting by conducting detailed analyses of each company's business environment and developing optimized cloud adoption strategies. Participating SMEs will be able to reduce consulting costs through government support while using the recommendations to accelerate AI transformation (AX) initiatives and improve business operations.

The company said it plans to leverage the project to expand customized technical support and advanced consulting services for enterprises preparing to adopt AI-driven digital transformation, further strengthening its cloud and AI business.

Lenovo Korea has completed the deployment of a Lenovo ThinkAgile HX hyperconverged infrastructure (HCI) solution powered by Nutanix at Jangan University, helping the institution modernize its IT infrastructure while improving service continuity.

Jangan University needed a highly reliable and easy-to-manage platform capable of delivering uninterrupted digital services to students and faculty despite having limited IT personnel. To replace its aging infrastructure, the university deployed a Nutanix HCI solution built on Lenovo ThinkAgile HX650 V3 nodes.

According to Lenovo Korea, the infrastructure modernization has significantly improved system uptime while reducing unexpected service interruptions, minimizing disruptions to the university's academic administration systems. The deployment has also reduced platform management and administrative workloads by approximately 30%, allowing IT staff to focus on more strategic initiatives.

The Lenovo ThinkAgile HX650 V3 integrates computing, storage and virtualization into a single platform and is powered by 5th Gen Intel Xeon Scalable processors. Lenovo said the system's highly scalable architecture enables organizations to expand infrastructure without disrupting existing operations, making it well suited for future growth.

ARGOS Identity, an AI-powered identity platform company, has been selected for the 2026 Future Industry-Based Enterprise Activation Support Program (Open Innovation) jointly operated by Gangnam-gu and The Invention Lab.

The program is an open innovation initiative designed to foster collaboration between large and mid-sized enterprises and startups. It supports proof-of-concept (PoC) projects and commercialization through business diagnostics, key performance indicator (KPI) development, one-on-one matching with corporate partners, investor presentations and demo days. A total of 108 companies applied, with eight ultimately selected.

Through the program, ARGOS Identity plans to expand the commercial adoption of Omni, its AI agent-based operations automation platform, by validating its use across multiple industries, including telecommunications and retail.

In the telecommunications sector, the company aims to automate document review during customer subscription and service activation. In retail, it plans to standardize unstructured data collected from physical stores and online channels before integrating the information into enterprise systems. In financial services, ARGOS Identity will explore automating Know Your Business (KYB) and Anti-Money Laundering (AML) review processes for global corporate customers.

According to the company, Omni automates repetitive document review and data verification tasks using AI agents. The platform analyzes external documents—including business registration certificates, contracts, application forms and supporting documents—verifies them against enterprise-defined policies, and transfers validated information into internal systems. Cases requiring additional review are routed to human operators through a human-in-the-loop (HITL) workflow.

ARGOS Identity said it has strengthened its AI-powered document analysis and identity verification technologies by supplying electronic Know Your Customer (eKYC) solutions to customers across financial services, fintech, commerce and ticketing industries. Through the new program, the company plans to expand beyond identity verification into AI agent-driven enterprise document automation.