Claude Opus 4.7 Now Generally Available
On April 16, Anthropic launched general availability of its latest flagship model, Claude Opus 4.7. The model delivers major improvements in reasoning and instruction following over Opus 4.6, with performance gains said to be especially notable on complex tasks.
The following day, April 17, Anthropic Labs also unveiled a new product called Claude Design. The tool lets users carry out design work in collaboration with Claude, showing Anthropic's product lineup expanding into support for creative work.
OpenAI and Microsoft Revise Their Partnership Agreement
OpenAI has revised its partnership agreement with Microsoft, its largest shareholder. Under the new agreement, OpenAI will be able to sell its products on Amazon Web Services (AWS), while Microsoft will gain more cash through a revenue-sharing arrangement. The deal removes a legal obstacle to OpenAI's $50 billion-scale Amazon transaction.
It marks a significant turning point, suggesting that the relationship between the two companies is shifting from its traditional exclusive partnership toward a more flexible one.
Ineffable Intelligence Raises $1.1 Billion
Ineffable Intelligence, a UK-based AI lab founded by former DeepMind researcher David Silver, has completed a $1.1 billion (approximately 165 billion yen) funding round just months after its founding. The company's valuation has reached $5.1 billion.
The company's goal is to build AI that learns without relying on human data. Silver is known for the self-play learning behind AlphaGo and AlphaZero, and expectations that autonomous learning through reinforcement learning will be key to next-generation AI appear to have attracted investors.
China Blocks Meta's Acquisition of Manus
Chinese authorities have blocked Meta's approximately $2 billion (roughly 300 billion yen) acquisition of Manus, an AI agent company, after a months-long investigation. Meta was ordered to cancel the acquisition, which could prove a setback for CEO Zuckerberg's push into the AI agent space.
Amid rising geopolitical tensions over AI technology, China's independent regulatory decision has had a direct impact on the strategy of a US company.
Anthropic and Amazon Sign Computing Deal for Up to 5 GW
On April 20, Anthropic announced an expanded partnership with Amazon, signing a new agreement to secure up to 5 gigawatts of computing resources. With securing the compute needed for AI model training and inference becoming a competitive point for companies, the deal is significant as a large-scale infrastructure investment.
Anthropic had also announced a next-generation TPU partnership with Google and Broadcom on April 6, pursuing a distributed infrastructure strategy built on multiple cloud providers.
Additionally, on April 24, the company announced that it will build one of Japan's largest AI-native engineering organizations in collaboration with NEC. Its expansion into the Japanese market is expected to begin in earnest.
Rumors of an OpenAI Smartphone in Development
According to a note from industry analyst Ming-Chi Kuo, OpenAI may be working with MediaTek, Qualcomm, and Luxshare to develop a smartphone. The concept reportedly involves an approach that replaces traditional apps with AI agents. However, there has been no official announcement at this point, and many details remain unclear.
---