2026.04.28 AI News Briefing — Microsoft–OpenAI exclusivity ends, Google DeepMind signs Korea MOU the same day

Today shook the global AI partnership map. Microsoft and OpenAI unwound their exclusive deal, making the multicloud era official, while Google DeepMind signed a 'K-Moonshot' cooperation MOU with the Korean government in Seoul. The same day, Anthropic's Claude took the wheel for NASA's autonomous drive on Mars.

---

1. Microsoft–OpenAI end their exclusive partnership — the multicloud era becomes official

• Source: OpenAI / Microsoft / CNBC / TechCrunch | 2026.04.27
• What happened: OpenAI and Microsoft substantially rewrote the exclusive partnership that had run for nearly seven years. Microsoft keeps its 'primary cloud partner' status, but OpenAI can now serve every product on any cloud. Microsoft's license to OpenAI IP runs through 2032 but is now non-exclusive, and Microsoft's obligation to share OpenAI-related revenue has ended. Revenue sharing from OpenAI to Microsoft continues through 2030, but with a cap on the total.
• Why it matters: Replacing the vague 'terminates automatically upon AGI' clause with a 2032 expiry removes the legal risk of a collision with OpenAI's $50 billion Amazon deal. It is a turning point that structurally locks in the multicloud shift in the enterprise AI market.

2. Google DeepMind signs MOU with Korea's Ministry of Science and ICT — 'K-Moonshot' accelerates

• Source: VOA Korea / Ministry of Science and ICT | 2026.04.27
• What happened: Google DeepMind signed a memorandum of understanding with Korea's Ministry of Science and ICT in Seoul. The two will pursue joint research in life sciences, weather and climate, and AI research tools, and expand an AI campus in Korea along with training and internship programs for researchers and students. Cooperation will center on the National Science AI Research Center due to launch this year. DeepMind CEO Demis Hassabis cited the tenth anniversary of the AlphaGo match and expressed a commitment to long-term collaboration with Korea.
• Why it matters: The government's AI-driven national R&D program, 'K-Moonshot', is now directly connected to a top-tier global AI lab. Visible gains in research productivity are expected in Korea's biotech and climate fields.

3. NASA's Perseverance makes its first 'AI-driven drive' — powered by Anthropic's Claude

• Source: Crescendo AI News | 2026.04
• What happened: NASA's Mars rover Perseverance completed a drive planned by AI for the first time ever. Anthropic's Claude vision-language model analyzed orbital imagery and terrain data to autonomously generate safe waypoints. Across two drives it covered a total of 456 meters, with AI taking over the complex planning work that human operators had done by hand for 28 years. NASA called it a 'big step' toward kilometer-scale autonomous exploration.
• Why it matters: It is a proven case of autonomous AI decision-making in space exploration, showing autonomous agents becoming mission operators in environments with long communication delays.

4. NVIDIA steps up support for Korea's AI ecosystem — releases Nemotron-Personas-Korea

• Source: NVIDIA Blog Korea | 2026.04.21–22
• What happened: Wrapping up 'Nemotron Developer Days Seoul 2026', NVIDIA released Nemotron-Personas-Korea, a Korea-specific synthetic dataset. It contains 6 million records built on official data from sources including the Korean Statistical Information Service (KOSIS), the Supreme Court, the National Health Insurance Service, the Korea Rural Economic Institute, and Naver Cloud. About 500 Korean AI developers and researchers attended, and a 'K-AI Builder Connect' session backed by the Ministry of Science and ICT and NIPA was held as well.
• Why it matters: With a synthetic dataset tailored to Korean industry and public-sector data now public, the barrier for Korean companies to fine-tune LLMs has dropped a notch.

5. Gartner: "AI-successful companies invest 4x more in data foundations" — only 39% confident

• Source: Gartner | 2026.04.16–28
• What happened: Gartner kicks off its global Data & Analytics Summit series on April 28–29, starting in São Paulo. Its published survey found that organizations succeeding with AI invest up to four times more, as a share of revenue, in 'foundational areas' such as data quality, governance, talent, and change management. Yet only 39% of the 353 global data and AI leaders surveyed said their current AI investments will contribute positively to financial results. Gartner recommends redesigning the 'context layer' as core infrastructure for AI agents.
• Why it matters: The data confirms that the gap in investment in foundational data and context, not the model, decides AI ROI. That carries a strong message for Korean companies stuck at the PoC stage.

6. Snap announces AI-driven mass layoffs — "65% of new code is written by AI"

• Source: Crescendo AI News | 2026.04
• What happened: Snap CEO Evan Spiegel announced roughly 1,000 layoffs and the closure of about 300 open positions. The company said that 'rapid advances in AI let us achieve the same results with fewer people', noting that more than 65% of new code is AI-generated. It expects annual savings of over $500 million by the second half of 2026, and the stock rose 11% in after-hours trading right after the announcement.
• Why it matters: It is a landmark case of a company explicitly citing 'AI replacing people' as the official reason. Pressure to redesign roles is now building across the tech industry.

---

Today's insight

• AI partnerships are losing their 'exclusivity': The Microsoft–OpenAI renegotiation signals the end of the 'single cloud + single model vendor' structure. Korean companies should now treat a multicloud, multi-model strategy as the default option to evaluate.
• Korea as a hub in the sovereign AI race: The Google DeepMind MOU and NVIDIA's Nemotron-Personas-Korea release landing on the same day is no coincidence. Korea is emerging as the 'Asian R&D and data hub' for global AI giants.
• Foundations, not models, decide ROI: As the Gartner survey shows, in an era where model performance gaps have narrowed, data quality, governance, and context infrastructure are the new competitive edge. Content governance domains like AEM and DAM are emerging as direct beneficiaries of this trend.