Google’s parent company, Alphabet, reported double-digit revenue growth for its second-quarter earnings on Wednesday. This is the tech giant’s 12th consecutive quarter of massive revenue gains.
The better-than-expected results come as investors evaluate how Alphabet’s huge spending on AI has affected its bottom line and what its long-term plans are to sustain those enormous investments. The company revised its yearly capital expenditure forecast to $200bn, with quarterly spending growing 100% year over year.
Alphabet is already one of the top five most valuable companies in the world, worth about $4.2tn. On Wednesday, it reported earnings of $9.11 per share, beating Wall Street’s expectations. Alphabet also surpassed revenue forecasting, coming in at $119.8bn. This brings the company roughly in line with its earnings during the first quarter of this year, which also saw a rise in revenue.
“Our AI investments are redefining what’s possible across every part of our business,” Sundar Pichai, the CEO of Google and Alphabet, said in a statement.
The gargantuan earnings come, however, as Google is recovering from delays in its most prominent and powerful AI model, Gemini Pro. The company has been a frontrunner in the AI race, but competitors, such as OpenAI and Anthropic, are neck-and-neck with the tech giant, which has prompted analysts to question whether Google can keep up.
On Tuesday, Google announced that it was releasing three cheaper AI models into its Gemini suite, including Gemini 3.6 Flash, Gemini 3.5 Flash Cyber and Gemini 3.5 Flash-Lite. The Flash Cyber model is geared toward cybersecurity and being able to patch security vulnerabilities, a capability that Google has been seen as lagging compared with its rivals.
Cybersecurity has become a major focus for AI models. But the technology also has drawbacks, including being vulnerable to exploitation by bad actors. Anthropic and OpenAI have been developing powerful models that experts warn would be disastrous if they became available to hackers.
Late on Tuesday, OpenAI revealed an “unprecedented cyber incident” in which its own AI agent went rogue and hacked a prominent startup.
Google said it was taking an “intentional approach” in deploying 3.5 Flash Cyber and it “will be exclusively available to governments and trusted partners”.
Cheaper Chinese AI models are also threatening Google’s top position in the AI battle, including Alibaba’s Qwen and a model called Kimi K3 by the startup Moonshot.
While demand has soared for Kimi K3, it has not caused the same amount of Wall Street panic seen in the deployment of other Chinese models, like DeepSeek.
