Kakao's turnaround hits roadblocks as AI, labor woes deepen

A person walks by Kakao's office in Pangyo, Gyeonggi Province, June 29. Yonhap

Kakao’s turnaround efforts are facing mounting challenges as weak artificial intelligence (AI) execution, slowing core businesses and escalating labor tensions cloud its growth outlook, even as it is expected to post another quarter of strong profits driven by restructuring and cost controls.

Samsung Securities said on Monday that Kakao’s second-quarter operating profit is expected to rise 18 percent year-over-year to 231.1 billion won ($154.4 million), while revenue is forecast to increase 1.2 percent to 2.48 trillion won, driven by asset sales and restructuring. However, the brokerage cut its target price to 44,000 won from 60,000 won, citing concerns that the company's growth strategy had hit roadblocks.

“Kakao’s introduction of AI agents, which it has positioned as a growth driver, is failing to deliver results, while its restructuring efforts are also losing momentum due to labor opposition," Samsung Securities analyst Oh Dong-hwan said in a report.

He added that it was difficult to justify a valuation premium over peers under the current circumstances, citing prolonged labor disputes and the lack of visible progress in AI commercialization.

In the past two years, Kakao has been slimming down its sprawling business portfolio under CEO Chung Shin-a, cutting the number of affiliates from over 140 to around 90 through mergers, divestments and the shutdown of noncore businesses.

The restructuring has helped the company post a record first-quarter operating profit of 211.4 billion won this year, with market consensus pointing to another improvement in the second quarter.

However, markets remain unconvinced of Kakao’s long-term growth prospects, viewing its earnings as driven primarily by cost cuts rather than improvements to underlying business fundamentals.

Kakao union members stage a rally during a partial strike in Seongnam, Gyeonggi Province, June 10. Yonhap

Kakao’s core businesses are showing signs of slowing momentum. Growth in its KakaoTalk advertising segment has eased, while commerce revenue has remained nearly flat despite rising transaction volumes. Its content division, including Japanese webtoon platform Piccoma, has contracted amid weaker demand and a lack of blockbuster titles.

More critically, the company’s biggest challenge remains its stalled AI strategy. Despite partnering with OpenAI and rolling out AI features within its flagship messenger platform KakaoTalk, the company has yet to demonstrate a clear monetization strategy or meaningful differentiation.

“Kakao is executing a two-track AI agent strategy with (its proprietary model) Kanana and (OpenAI’s) ChatGPT, but both services are underperforming,” Oh said. “AI services need to be redesigned from the ground up from a consumer perspective rather than a supplier perspective.”

Risks have also increased as internal tensions deepen. Kakao’s labor union, including employees across five affiliates, has continued industrial action over compensation, wage increases and job security following restructuring. The union is scheduled to hold another protest on July 21, following strikes held last month.

Its founder Kim Beom-su’s ongoing legal proceedings over alleged stock price manipulation during Kakao’s acquisition of SM Entertainment have added another layer of governance uncertainty.

In contrast, its rival Naver has been pursuing an investment-heavy approach that prioritizes long-term growth over short-term profitability. The company is expanding into AI infrastructure and enterprise services, including plans to build gigawatt-scale AI data centers in partnership with Nvidia, even as rising costs weigh on margins.

Naver has flagged AI infrastructure as a key medium to long-term growth driver, unveiling plans last month to build an AI factory data center with a capacity of around 1 gigawatt over the next five to six years.

The company aims to generate about 20 trillion won in new annual revenue from the project and plans to take part in government-led megaprojects to build regional AI data centers across the country.