CMOs Double Down on GenAI: $ 10M+ Annual Investments on the Rise

Here is what’s new in the AI world.

71% of CMOs to Allocate Significant Budgets to GenAI Over Next Three Years”

A new report from Boston Consulting Group (BCG), titled "How CMOs Are Scaling GenAI in Turbulent Times", reveals that nearly 71% of Chief Marketing Officers (CMOs) plan to allocate over $10 million annually to generative AI (GenAI) initiatives over the next three years, a significant jump from 57% just a year ago.

The findings come from BCG’s annual global survey of 200 CMOs across Asia, Europe, and North America, conducted in April–May 2025. Optimism around GenAI continues to rise among CMOs, growing from 74% in 2023 to 83% in 2025, while concerns have markedly declined.

GenAI’s Strategic Role Expands

According to Parul Bajaj, who leads BCG’s marketing, sales, and pricing practice in India, GenAI’s most profound impact is in delivering personalization at scale, speeding up content creation, particularly in immersive formats like video, and transforming the entire customer journey.

She noted that B2B marketers are especially ahead, embracing AI agents, dynamic product descriptions, and predictive analytics to weave GenAI into both strategy and execution.

Though a third of CMOs have seen improvements in customer experience and content quality due to GenAI, fewer reported gains in efficiency compared to previous years. Still, nearly 60% of CMOs anticipate at least a 5% increase in revenue in their targeted areas through GenAI adoption.

Personalization and Content Take the Lead

While GenAI has enhanced content volume, quality, and customer experience for over one-third of CMOs, fewer now cite productivity improvements or reductions in manual tasks.

BCG’s Mark Abraham, a global personalization leader, said GenAI is becoming a core part of marketing, with CMOs increasingly moving from isolated pilot programs to broader implementation, despite economic uncertainty.

Areas like immersive media, video generation, AI-driven personalization, and agent-based AI are emerging as top investment priorities. Notably, 30% of CMOs identified video generation as their next major focus in content creation.

From Pilots to Scaling Impact

CMOs reported that some use cases like product recommendations, content timing, and selecting next-best content have already been fully scaled. Other use cases, such as personalized offers, churn prediction, and audience targeting, are still in pilot stages. Some marketers shared that scaling personalized offers resulted in returns three times greater than those from generic campaigns.

AI agents are gaining traction, especially in B2B sectors. About one in three B2B CMOs and one in four B2C CMOs listed AI agents as a top investment area for the future of marketing workflows.

Measurement and Talent Still Lagging

Interestingly, using GenAI to track marketing ROI ranks lowest among investment priorities. Many CMOs are still relying on earlier investments in first-party data and have yet to significantly improve their measurement capabilities.

With GenAI talent in short supply, companies are prioritizing internal upskilling. Rather than hiring externally, many are turning to hands-on training programs such as hackathons, AI incubators, and internal demos to build AI fluency within their teams.

Lauren Wiener, a global lead in BCG’s marketing division, emphasized that the most successful CMOs are not just adopting GenAI tools—they’re also investing in their people. “The leaders are the ones who scale both technology and talent,” she said.

Meta Unveils GenAI Plans to Fully Automate Ad Creation by 2026

Meta Platforms aims to launch AI tools capable of creating and targeting entire advertisements by late 2026, according to a Wall Street Journal report. This initiative is central to CEO Mark Zuckerberg's strategy for Meta's future, where advertising remains its primary revenue source.

The technology could particularly benefit small and medium-sized businesses (SMBs), a major segment of Meta's advertisers, who often lack resources for complex ad creation.

However, larger retail brands reportedly expressed reservations about relinquishing creative control, citing concerns that AI might not consistently match human-made campaign quality or aesthetics, sometimes producing visuals needing refinement.

This represents a significant expansion beyond Meta's current AI tools, which mainly generate variations of existing ads. The envisioned workflow would let brands submit a product image and budget goal; the AI would then produce complete ads (images, video, text), determine optimal targeting for Facebook/Instagram users, and suggest budget allocations.

Meta also plans real-time AI personalization, tailoring ad versions based on factors like user location. Meta enters a competitive AI content creation space where Google offers tools like Veo, and brands frequently use third-party platforms (e.g., Midjourney, DALL-E). Meta is exploring ways to integrate such third-party capabilities.

The push follows a recent reorganization of Meta's AI teams into two groups: one focused on product development and another on foundational technology. This restructuring aims to streamline operations amid reported challenges in retaining top AI talent, with several researchers having joined French startup Mistral.

AI as a Scientist? Sam Altman Says the Future Is Closer Than You Think

At the Snowflake Summit 2025, OpenAI CEO Sam Altman shared his vision of artificial intelligence evolving beyond its current role as a simple assistant. According to Altman, AI is transitioning from being a tool to acting as a teammate, capable of handling tasks, improving through feedback, and even potentially generating new knowledge.

Altman likened the way people now interact with AI tools like ChatGPT to managing junior employees. Users are assigning tasks to AI agents, reviewing their output, giving guidance, and iterating, just as they would with entry-level team members.

“You’ll hear people say their job now involves directing agents, reviewing their work, providing feedback—it’s almost identical to working with a team of junior staff,” Altman said.

Looking ahead, Altman predicted that by 2026, AI agents may go beyond assistance and begin making novel discoveries or solving complex business challenges. “I’d wager that next year, in some limited ways, we’ll see agents uncover new knowledge or find non-obvious solutions to real problems,” he added.

Former OpenAI Scientist Sought a Bunker Before AGI Launch

In related news, former OpenAI co-founder and chief scientist Ilya Sutskever reportedly told his team in 2023 that the company would need to build a secure bunker before deploying artificial general intelligence (AGI).

“We’re definitely going to build a bunker before we release AGI,” Sutskever said during an internal meeting, according to a new book, Empire of AI by journalist Karen Hao. He later clarified that access to the bunker would be optional.

The revelation is one of several disclosed in the book, which chronicles the intense internal struggles within OpenAI, including the brief removal of Sam Altman as CEO in late 2023, a power struggle in which Sutskever played a key role before ultimately leaving the company.

U.S. Leverages AI Diplomacy Through OpenAI’s Global Vision

OpenAI has launched a new global initiative, “OpenAI for countries,” offering what it brands as AI infrastructure aligned with democratic values. This initiative, marketed as “AI on democratic rails,” is both a strategic political move and a major business opportunity.

Under the plan, OpenAI would control the entire AI stack, spanning from data centers and foundation models to user-facing applications. Elon Musk even attempted to position his own AI venture, xAI, as part of the deal, indicating the high stakes.

But what does AI on "democratic rails" actually mean? According to OpenAI’s own blog, it implies AI that upholds democratic values and human rights, and prevents state misuse of the technology for authoritarian control. This pitch resonates with broader concerns over the geopolitical AI race between the U.S. and China.

However, OpenAI’s first partnership under this program was with the United Arab Emirates, a nation that Freedom House rates as “Not Free,” with well-documented human rights abuses including political repression and censorship. This casts doubt on OpenAI's ability to genuinely promote democratic principles in such environments.

Another promise OpenAI makes is to support national data sovereignty through the development of in-country data centers. But this clashes with the U.S. Cloud Act, which allows American law enforcement to demand access to data held by U.S. companies, even if that data resides overseas. Given OpenAI’s stated intention to collaborate closely with the U.S. government, assurances of sovereignty appear questionable.

In reality, countries that sign comprehensive agreements with OpenAI may end up exposing themselves to American oversight of their AI infrastructure. This would give the U.S. potential access to sensitive national data and undermine digital independence.

U.S. allies in Europe have grown increasingly wary of how American tech companies are used to further U.S. foreign policy objectives. A recent example involves the International Criminal Court’s chief prosecutor, Karim Khan, losing access to his Microsoft email account following U.S. sanctions under the Trump administration.

In another case, Ukrainian officials lost access to satellite data from Maxar, a U.S. company, amid political pressure. If the U.S. can already restrict access to critical technology, expanding that control to AI infrastructure would significantly increase its global leverage.

OpenAI’s timing appears more economically motivated than principled. Investor hesitation around its massive $500 billion Stargate project has pushed the company to seek foreign government deals as an alternative revenue. This strategy also aligns with U.S. goals to create jobs and extend its tech footprint globally, advancing both economic and strategic national interests, potentially at the cost of the sovereignty of partner nations.

As recent U.S. actions have shown, alliances with American firms can quickly become liabilities. Countries embracing “OpenAI for countries” risk losing in-house expertise and becoming reliant on U.S. technology, and, consequently, U.S. political whims. Despite its emphasis on democratic ideals and data autonomy, this initiative appears to prioritize American interests above all.

Your opinion matters!

Hope you loved reading our piece of newsletter as much as we had fun writing it. 

Share your experience and feedback with us below ‘cause we take your critique very critically. 

Thank you for reading

-Shen & Towards AGI team