Big Tech’s Q2 results: AI is moving from improving advertising to running it

Three months ago, Big Tech's first-quarter earnings delivered a relatively reassuring message for advertisers. Artificial intelligence may have dominated product launches, investor calls and capital expenditure plans, but advertising remained the business funding it all.

Read On: Big Tech Q1 results: AI may be the story, but ads are still the business

The second quarter suggests that relationship is evolving.

Alphabet, Meta, Amazon and Microsoft collectively reported another quarter of robust growth, driven by digital advertising, cloud computing and enterprise software. Alphabet's revenue rose 24% to $119.8 billion, Meta reported a 28% jump to $60.8 billion, Amazon crossed the $200 billion mark in quarterly revenue for the first time at $200.6 billion, while Microsoft posted revenue of $90 billion, up 18%.

Search, social advertising, commerce media and cloud businesses all continued to expand despite persistent macroeconomic uncertainty and questions around AI monetisation.

The infrastructure race, meanwhile, has only intensified. Alphabet expects to spend about $180-190 billion on capital expenditure this year, Meta has narrowed its outlook to $130-145 billion, Amazon has raised planned investment to roughly $220 billion and Microsoft continues investing aggressively in Azure infrastructure. Together, the four companies are preparing to spend well over $700 billion building the data centres, chips and computing capacity that will underpin the next phase of AI.

But for marketers, the more interesting shift lies elsewhere.

Three months ago, AI largely functioned as an optimisation layer. It improved targeting, recommendations and creative efficiency. This quarter, it is beginning to look more like an operating layer for advertising itself.

Read On: AI is rewriting Indian marketing's OS. Is the industry ready?

Across search, social media, commerce and enterprise software, the largest technology companies are steadily expanding AI's role from improving campaigns to deciding how campaigns are created, bought, delivered and measured.

Google and Meta: From better yield to greater control

Alphabet's results should have been awkward for anyone predicting AI would cannibalise Google's advertising engine.

Search and Other revenue grew 17% to $63.3 billion, YouTube advertising increased 13% to $11.1 billion, while overall Google Advertising revenue reached $81.6 billion. At the same time, Google Cloud revenue surged to $24.8 billion as enterprise AI demand accelerated.

The significance is not simply that Google's advertising business remains healthy.

It is that Search continues growing even as Google fundamentally changes the search experience through AI Overviews and AI Mode. The question has shifted from whether AI search can carry advertising to how Google monetises AI-generated answers without undermining user trust or blurring the distinction between recommendation and promotion.

For advertisers, this could gradually change what search optimisation means. Winning a keyword auction may become only part of the equation. Visibility inside AI-generated answers, product recommendations and shopping summaries may increasingly determine commercial success.

Meta tells a complementary story.

The company reported revenue of $60.8 billion, up 28%, while daily active people across its family of apps reached 3.6 billion. AI continues improving recommendations and advertising performance, helping keep engagement high across Facebook and Instagram even as Meta commits between $130 billion and $145 billion towards AI infrastructure this year.

Yet the larger shift is no longer about recommendation algorithms alone.

Meta increasingly wants advertisers to define objectives rather than campaigns. AI is steadily taking over creative generation, audience selection, budget allocation, placement optimisation and performance improvements through products such as Advantage+. The marketer supplies the goal. Increasingly, Meta supplies the execution.

Read On: Why India remains Meta’s most important unfinished market

In Q1, AI primarily improved advertising yield.

In Q2, Google and Meta are showing signs of redesigning different parts of the advertising value chain. Google is reshaping the surface on which advertising appears. Meta is reshaping the process through which advertising is bought.

Amazon: The closed loop gets wider

Amazon remains structurally different from the other platforms because its advertising business begins with transactions rather than attention.

That distinction became even more apparent this quarter.

Amazon reported revenue of $200.6 billion, while AWS grew 37% to $42.2 billion, its fastest growth in more than four years. Advertising revenue rose 26% to $19.8 billion, continuing to outpace much of the broader digital advertising market. The company also increased planned capital expenditure for the year to around $220 billion as demand for AI infrastructure continues to exceed available capacity.

For marketers, however, the more important development is not simply advertising growth.

Amazon is steadily expanding beyond sponsored product listings into a broader commerce media ecosystem spanning Prime Video, streaming inventory, off-site advertising and increasingly sophisticated measurement built around first-party commerce data.

AI strengthens that proposition by connecting retail signals across product discovery, creative generation, audience targeting and campaign optimisation while maintaining the closed-loop attribution that has become Amazon's biggest competitive advantage.

In effect, Amazon is extending the commerce signal beyond the marketplace itself.

Read On: Retail media grows from ad channel to digital shelf rent

That matters because commerce media is no longer simply a retailer selling advertising inventory. It is increasingly becoming a media ecosystem built around purchase intent.

Microsoft: AI enters the marketing organisation

Among the four companies, Microsoft remains the least dependent on advertising revenues. But its earnings may prove equally significant for marketers.

Revenue rose 18% to $90 billion, while Microsoft Cloud revenue reached $59.3 billion. Azure continued growing strongly, and search and news advertising also increased during the quarter. Copilot adoption across Microsoft 365 and enterprise software continued expanding as businesses integrated AI into everyday workflows.

Unlike Google, Meta and Amazon, Microsoft's primary role is not to automate media platforms.

It is increasingly automating the organisations buying media.

Research, planning, content generation, CRM, analytics and productivity are steadily becoming AI-assisted through Microsoft's software ecosystem. Azure provides the infrastructure. Copilot increasingly becomes the interface through which marketers interact with it.

Read On: When did CRM quietly become marketing?

That positions Microsoft differently from the advertising platforms.

Google, Meta and Amazon are embedding AI inside consumer experiences. Microsoft is embedding AI inside marketing departments.

India: Cheaper sophistication, deeper dependency

For Indian marketers, these shifts are likely to be felt less through spectacular new AI products than through the gradual automation of everyday marketing.

Performance Max, Advantage+, retail media platforms and AI-powered creative tools will continue lowering the operational complexity of campaign management. Smaller brands and regional advertisers will gain access to optimisation capabilities that previously required specialist teams, while vernacular creative generation and automated asset production become increasingly accessible.

The trade-off is that more campaign decisions will move inside platform models.

As Google, Meta and Amazon automate targeting, bidding, creative optimisation and measurement, marketers may achieve better outcomes while understanding less about why campaigns perform the way they do.

Read On: As clicks disappear, India’s SEO agencies rethink what they sell

That also raises the value of assets platforms cannot easily replicate.

First-party customer data, commerce signals, proprietary content and direct customer relationships become more valuable as AI systems take over more of the mechanics of advertising. Agencies, meanwhile, may increasingly compete on strategy, measurement and business understanding rather than campaign execution alone.

TL;DR: AI's next role in advertising

If Q1 established that AI was reinforcing the existing advertising model, Q2 suggests it is beginning to reorganise it.

Google is steadily controlling not just discovery but increasingly the answer itself. 

Meta is moving towards campaigns where marketers define objectives while AI handles execution. 

Amazon is widening the commercial signal from ecommerce into a broader commerce media ecosystem. 

Microsoft is becoming the software layer through which marketing organisations increasingly operate.

The common thread across all four companies is not simply bigger AI investments. It is that AI is moving beyond improving advertising performance to becoming the system through which advertising is conceived, bought, delivered and measured.

And that may prove a far more consequential change than any individual chatbot or model launch.