AI Sector Daily Digest — July 8, 2026

1. OpenAI prepares GPT-5.6 for public launch after U.S. review

• OpenAI plans to publicly launch GPT-5.6 on Thursday after delaying the release at the U.S. government's request. 1
• The rollout includes GPT-5.6 Sol and lower-cost Terra and Luna models; access had been limited to vetted partners. 1
• The launch path matters because frontier-model releases are starting to include a government review step, not just a product-readiness decision. 1

2. Beijing weighs limits on overseas access to advanced Chinese AI models

• Chinese authorities have discussed possible overseas-access limits with Alibaba, ByteDance and Z.ai, Reuters reported, citing three people familiar with the talks. 2
• The discussions cover both closed and more open model releases, plus tougher penalties for leaks or theft of proprietary AI technology. 2
• The scope is still unsettled, but the move would treat frontier AI models more like strategic technology than ordinary software exports. 2

3. DeepSeek is reportedly designing its own inference chip

• DeepSeek is developing an AI chip for inference, the stage when a trained model generates answers for users, Reuters reported, citing three people familiar with the matter. 3
• The effort is early: the company has been speaking with chip-design, foundry and memory partners, and has privately increased chip-engineering hiring. 3
• If it works, DeepSeek would reduce its dependence on Nvidia and Huawei hardware, but manufacturing and high-bandwidth memory restrictions remain major obstacles. 3

4. Norm Ai raises $120 million for AI-powered legal services

• Norm Ai raised $120 million in a new round that values the legal-AI startup at $1.2 billion. 4
• Khosla Ventures led the financing, with Blackstone, Bain Capital Ventures and Coatue among the participants. 4
• Norm is pairing attorneys with AI engineers, and it says the funding will support supervisory AI systems for regulated industries as well as hiring. 4

5. Bank of England warns AI can amplify market and cyber risk

• The Bank of England said AI poses a growing threat to financial stability because investors are making large bets on adoption while banks face higher cyber and operational exposure. 5
• The central bank warned that a reassessment of AI companies' earnings prospects could hit equity prices and worsen volatility if positions are crowded or leveraged. 5
• It also said more capable AI may force financial firms into more frequent software updates, which can improve defenses but also create operational-disruption risk. 5