AI News (2025/8/27): OpenAI Enhances ChatGPT Mental Health Support; Cambricon Revenue Soars…

AI Safety Upgraded: OpenAI Enhances Mental Health Support, Chinese AI Chip Manufacturer Revenue Soars

Summary

• OpenAI plans to update ChatGPT to better respond to users' mental health needs, provide parental controls, and enhance safety measures in conversations about suicide.
• Chinese AI chip design company Cambricon, benefiting from government policies encouraging the use of domestic technology, reported a 44-fold revenue increase in the first half of the year, with a profit of about $144 million.
• Anthropic has reached a settlement with authors over copyright issues in AI training data.
• Google DeepMind's weather lab has surpassed traditional models in hurricane prediction.
• Verily (Alphabet's life sciences division) is laying off employees and ending its medical device projects, shifting focus to precision medicine, data, and AI.
• Microsoft headquarters was temporarily locked down due to a protest.

Major News

1. OpenAI Enhances ChatGPT Mental Health Support: OpenAI announced plans to upgrade ChatGPT to better identify users expressing mental distress, provide more robust parental control features, and enhance safety measures in conversations about sensitive topics such as suicide. This move indicates that OpenAI is taking the potential risks of large language models seriously and is actively exploring responsible AI development paths (Source).
2. Cambricon Revenue Soars: Benefiting from the Chinese government's policies encouraging the use of domestic technology, Chinese AI chip design company Cambricon reported a 44-fold revenue increase in the first half of 2025, achieving a profit of about $144 million. This reflects the rapid improvement in China's self-research and development capabilities in the AI chip sector and the gradual reduction of dependence on foreign technology (Source).
3. Anthropic Reaches Copyright Settlement: AI company Anthropic has reached a settlement in a copyright class-action lawsuit. The lawsuit, initiated by authors, accused Anthropic of using a pirated database containing copyrighted works to train AI models. The specific terms of the settlement have not been disclosed, but it indicates that AI companies are becoming more aware of copyright issues and are taking steps to address related disputes (Source).
4. Google DeepMind Surpasses Traditional Models in Hurricane Prediction: Google DeepMind's weather lab has made significant progress in hurricane prediction. Data shows that its model is more accurate than traditional models in predicting the path of Hurricane Erin over the next 72 hours. This demonstrates the enormous potential of AI in meteorological prediction and its potential to help people better cope with extreme weather events (Source).
5. Alphabet's Verily Department Adjusts Strategy: Alphabet's life sciences division, Verily, announced layoffs and the termination of its medical device projects, focusing instead on precision medicine, data, and AI. This strategic adjustment reflects Verily's exploration direction in healthcare and the significant potential of AI and data-driven precision medicine (Source).
6. Microsoft Headquarters Temporarily Locked Down Due to Protest: A protest at Microsoft headquarters saw demonstrators entering President Brad Smith's office, demanding that Microsoft sever ties with the Israeli government. Microsoft subsequently temporarily locked down the building and held an emergency press conference. This incident highlights the growing pressure on tech companies in geopolitical and social issues (Source).

Quick Updates

• Anthropic has released Claude for Chrome, allowing Claude to perform actions on behalf of users in the browser. It is currently available as a research preview to 1,000 Max subscribers (Source).
• Indian IT services giant TCS has established a new AI operations department and appointed Amit Kapur as its head (Source).
• Trump stated that Meta plans to invest $50 billion in the "Hyperion" data center under construction in Louisiana, while Meta previously indicated an investment exceeding $1 billion (Source).
• Blue Water Autonomy, which builds autonomous unmanned ships for the U.S. Navy, has completed a $50 million Series A funding round led by GV, bringing its total funding to $64 million (Source).
• Okta released its Q2 earnings report, showing a 13% year-over-year revenue increase to $728 million, with net income growing 131% to $67 million and an increased outlook for fiscal year 2026 (Source).
• MongoDB released its Q2 earnings report, showing a 24% year-over-year revenue increase to $591.4 million, with net loss decreasing 14% to $47 million and an improved full-year guidance (Source).
• Framework is developing a large haptic touchpad, Trackpoint pointing stick, and an eGPU for its laptops (Source).
• Rumors suggest that Sequoia China (now HongShan) invested only a quarter of the approximately $9 billion it raised in 2022 and is increasingly looking at deals outside of China (Source).

Market/Community Trends

AI technology is expanding into a wider range of applications. Beyond traditional business uses, AI is showing significant potential in areas such as mental health support and weather prediction. At the same time, competition in the AI sector is intensifying, with Chinese companies accelerating their rise and gradually reducing their dependence on foreign technology. Ethical and safety issues in AI are receiving increasing attention, and companies like OpenAI are actively exploring responsible AI development paths. Additionally, the emergence of AI super PACs indicates that AI is influencing politics, which may spark ethical discussions about its application in this domain (Source). In Asia, AI companies are actively forming partnerships and providing free AI tools to users to gather more real-world consumer data for training (Source).

---