OpenAI CEO Sam Altman has told employees the company is open to slowing down the development of its AI systems. This shift addresses growing safety concerns from researchers like Jacob Coxon, who warned that the industry is racing too fast toward self-improving superintelligence without adequate safeguards.
Following Coxon's resignation from Anthropic and his viral warnings about AI risks, Anthropic announced it has added stronger safeguards to block malicious activities. Meanwhile, the global AI security market is expected to reach $4.78 billion by 2025, driven by a 68.7% increase in demand for protection against cyber threats.
In the legal sector, Latham & Watkins purchased Nvidia GPU servers to build internal AI models, ensuring sensitive client data remains secure outside of cloud vendors. Similarly, Akerman hired its first Director of AI Development to guide the firm's adoption of artificial intelligence tools.
Google has delayed the launch of its Gemini 3.5 Pro model, which remains under testing with partners as of September 10. This follows a leadership change at Google DeepMind, where Zoubin Ghahramani replaced Demis Hassabis in August.
While AI continues to transform financial trading with faster data analysis and automation, the sector faces scrutiny. SB Energy's IPO details reveal significant risks, noting a $3.2 billion loss in the first half of 2026 and a controversial deal involving $5.5 billion in warrants to secure data center leases from OpenAI.
Key Takeaways
- OpenAI CEO Sam Altman told staff the company may slow down AI development to address safety concerns.
- Anthropic researcher Jacob Coxon resigned and warned that AI companies are prioritizing speed over safety.
- Anthropic announced new safeguards to block malicious activities like cyberattacks and biological weapon research.
- The global AI security market is projected to reach $4.78 billion by 2025.
- Google has delayed the release of its Gemini 3.5 Pro model despite earlier promises.
- Latham & Watkins bought Nvidia GPU servers to keep client data secure within its own systems.
- Akerman hired Charles Zerner as its first Director of AI Development to build internal AI solutions.
- AI systems recently solved a famous Millennium Prize problem in mathematics.
- SB Energy lost $3.2 billion in the first half of 2026 while seeking OpenAI data center leases.
- SB Energy offered warrants worth $5.5 billion to secure OpenAI's data center support.
OpenAI CEO Says Company May Slow AI Development
OpenAI CEO Sam Altman told employees that the company is open to slowing down the development of its AI systems. This announcement comes amid growing safety concerns from researchers like Jacob Coxon, who warned that AI companies are racing too fast. Altman suggested that OpenAI might coordinate with other AI labs to pace their advancements more carefully. The company previously noted in July that the speed of AI growth might eventually require a global slowdown.
Anthropic Researcher Jacob Coxon Resigns Over AI Safety
Jacob Coxon, a researcher at Anthropic and former OpenAI employee, resigned and publicly warned that AI companies are prioritizing speed over safety. He stated that both Anthropic and OpenAI are racing toward self-improving superintelligence without adequate safety measures. His post went viral with over 100 million views and highlighted fears that advanced AI could pose a serious threat to humanity by the end of the decade. In response, Anthropic announced it has added stronger safeguards to block malicious activities like cyberattacks and biological weapon research.
OpenAI CEO Considers Slowing Pace of AI Development
A report by The Information reveals that OpenAI CEO Sam Altman told staff the company is willing to slow down cutting-edge AI development. Altman expressed this willingness during a private meeting to address concerns about the technology's safety and ethics. This represents a significant shift from OpenAI's previous focus on pushing the boundaries of what is possible with AI. The company is currently discussing the issue internally while facing scrutiny from experts and investors.
Indian Retail Traders Gain Access to AI-Powered Algo Trading
Retail algorithmic trading in India is becoming more accessible due to new broker APIs, cloud infrastructure, and AI-assisted coding tools. These technologies lower the cost and technical barriers that previously limited algorithmic trading to large institutions. Experts note that while tools like drag-and-drop platforms make coding easier, traders still need robust strategies and risk management. Regulatory clarity is also helping to formalize this growing ecosystem for individual investors.
AI Transforms Financial Trading With Faster Data and Automation
Artificial intelligence is changing financial markets by allowing traders to analyze huge amounts of data in seconds instead of hours. AI systems can identify market patterns and automate repetitive tasks, helping investors make more data-driven decisions. Automated trading allows software to execute orders based on specific rules, which reduces the impact of human emotions like fear or greed. However, experts warn that AI cannot guarantee profits and traders must still carefully test their strategies and manage risks.
Ex-Anthropic researcher Jacob Coxon warns AI risks
Jacob Coxon resigned from his job at Anthropic on Tuesday. He publicly accused the company and OpenAI of creating AI systems that could threaten humanity. Coxon shared his concerns in an extended interview about the dangers of artificial intelligence.
AI security market grows fast as risks rise
The global AI security market is expected to reach $4.78 billion by 2025. This represents a 68.7% increase driven by the need to protect against cyber threats. NVIDIA CEO Jensen Huang noted that interest in AI-powered security solutions is growing rapidly. The market is projected to grow at a 24.3% annual rate from 2022 to 2025.
Google delays launch of Gemini 3.5 Pro model
Google has not yet released its Gemini 3.5 Pro AI model despite earlier promises. CEO Sundar Pichai mentioned the model in July, but it is still not available as of September 10. Google DeepMind states the model is currently being tested with partners. Meanwhile, Google replaced long-time leader Demis Hassabis with Zoubin Ghahramani in August.
Latham law firm buys Nvidia servers for AI
Latham & Watkins, a major US law firm, purchased Nvidia GPU servers to build its own AI models. The firm spent heavily on this infrastructure to keep sensitive client data secure within its own systems. Rene Mendoza, the chief information officer, explained that they avoid cloud vendors to protect confidential information. This approach differs from other firms that rely on external platforms like Harvey and Legora.
Open Secure AI Alliance joins Linux Foundation
The Open Secure AI Alliance has moved its home to the Linux Foundation for neutral governance. This change allows the group to develop open AI security tools without influence from any single company. The alliance aims to create trustworthy AI systems for organizations of all sizes. The Linux Foundation will support their mission to advance secure AI technologies.
Akerman hires first director for AI development
Akerman has hired Charles Zerner as its first Director of AI Development. Zerner will use his experience in technology and law to help lawyers choose the right AI tools. He will also work on building new AI solutions inside the firm. This move shows that Akerman is taking a serious step toward using artificial intelligence in legal work.
AI solves Millennium Prize problem in math
Artificial intelligence recently solved a famous math puzzle known as the Millennium Prize problem. This achievement proves that AI systems can now handle complex tasks that were once thought impossible. The solution combines human creativity with AI power and marks a major milestone in mathematics. Experts say this result ends the joke that AI cannot even do simple calculations.
SB Energy IPO reveals warning signs for AI investors
SB Energy plans to go public but its IPO details warn investors about potential risks in the AI market. To get OpenAI to lease its data centers, SB Energy gave away warrants worth $5.5 billion. These warrants allow OpenAI to buy shares at a very low price if certain conditions are met. The company lost $3.2 billion in the first half of 2026 and has few completed data centers. This deal suggests that demand for AI data centers may not be as strong as some believe.
Sources
- Altman tells staff OpenAI is open to slowing AI development, Bloomberg News reports
- Anthropic researcher Jacob Coxon quits over AI safety: What he warned and how OpenAI, Anthropic reacted
- OpenAI open to slowing cutting-edge AI, CEO Sam Altman tells staff
- Retail algo trading gets a makeover: how APIs, AI and regulation are opening the door for Indian traders
- How AI Is Changing the Way People Trade in Financial Markets
- Extended Interview: Ex-Anthropic researcher Jacob Coxon, who warns AI could destroy humanity
- AI Security Industry Booms as Risks Escalate
- Where is Gemini 3.5 Pro? The case of the missing AI model.
- Latham builds its own AI models with Nvidia GPU server investment
- Open Secure AI Alliance Finds a Neutral Home at the Linux Foundation
- Akerman Hires First Director of AI Development
- Millenium Prize solution shows the case for artificial intelligence is adding up
- SB Energy IPO: This Hidden Detail Is a Major AI Bubble Warning
Comments
Please log in to post a comment.