SK Telecom's AI Data Center MOU Reinforces Growth Ambitions Amid Unresolved Financial Pressures

What happened

SK Telecom has signed a non-binding memorandum of understanding with Supermicro and Schneider Electric to develop pre-fabricated modular solutions for AI data centers, aiming to accelerate deployment and reduce costs. This move aligns with SKM's existing AI infrastructure scale-up, highlighted in the DeepValue report, which notes FY2025 AI data center revenue grew 34.9% year-over-year to KRW 519.9 billion. However, the report critically emphasizes that SKM's near-term equity story is dominated by cyber liability risks and dividend uncertainty, not by AI asset monetization which remains accounting-driven. The company faces a potential KRW 2.3 trillion compensation payout from a consumer agency recommendation and has canceled FY2025 dividends, conditioning future returns on improved cash flow. Thus, while the MOU supports long-term growth narratives, it does little to address immediate financial overhangs or provide a clear path to shareholder value creation.

Implication

For investors, this announcement reinforces SKM's strategic focus on AI data centers, potentially enhancing future revenue growth as global demand rises. Yet, the DeepValue report underscores that AI-related gains are primarily fair-value marks without disclosed cash monetization, limiting near-term financial impact. SKM's elevated leverage—net debt-to-EBITDA of 8.04—and unresolved compensation liabilities, estimated at KRW 2.3 trillion, threaten cash flow and balance sheet flexibility. Dividend cancellations have erased the traditional telecom payout anchor, making returns contingent on uncertain AI optionality rather than stable income. Therefore, investors should view this news as incremental to the growth story but insufficient to alter the risk-reward profile without progress on bounding liabilities and reinstating shareholder returns.

Thesis delta

The MOU does not materially shift the investment thesis, which remains dependent on bounding cyber liabilities and reinstating a quantified payout policy, as outlined in the DeepValue report. It may slightly strengthen the bull case by supporting AI infrastructure growth, but without addressing monetization or cash return mechanisms, the core thesis unchanged requires monitoring for execution on liabilities and dividends over the next 3-6 months.