SMX (Security Matters) is preparing to launch its AI-powered platform, SMX Scientific Intelligence, internally in Q1 2027, with full operation and a limited commercial version planned for Q2 2027. The platform aims to accelerate material research by combining AI with scientific data to identify chemical agents, additives, and molecular markers. It will also support recycled plastic certification and the Plastic Circularity Credits marketplace, though certified recycled-plastics standards have not yet been adopted.
In the banking sector, a McKinsey report indicates AI could eliminate 60 to 70 percent of manual tasks in risk functions. A survey of 300 risk and treasury leaders across 31 countries found that 59 percent of banks have embedded AI in decision-making, but only 18 percent have fully operational AI governance. Most banks still restrict autonomous AI decisions, though 76 percent expect scaled use by 2031.
Red Hat is developing security tools for AI agents that take actions rather than just answering questions. Its OpenShell runtime provides agents with masked credentials so they never hold real keys, and agents can run inside Kata Containers or virtual machines on OpenShift. Red Hat is collaborating with OpenAI Group on standards for agent identity and authorization.
Key Takeaways
- <ul><li>SMX plans internal launch of Scientific Intelligence AI platform in Q1 2027, with full operation and limited commercial version in Q2 2027</li><li>AI could remove 60-70% of manual bank risk tasks, per McKinsey</li><li>Only 18% of banks have fully operational AI governance despite 59% embedding AI in decision-making</li><li>Red Hat's OpenShell runtime gives AI agents masked credentials and runs them in containers or VMs on OpenShift</li><li>Red Hat is working with OpenAI Group on agent identity and authorization standards</li><li>AST SpaceMobile targets $1 billion revenue post-commercial launch and partners with Google and Meta on satellite-based AI</li><li>Crusoe sold unfinished Wyoming data center to Google for $17 billion buildout</li><li>Alterra Labs raised $12.5 million seed round to train medical AI in simulated environments</li><li>2CRSi partners with UET to deploy immersion-cooled AI infrastructure for military and industrial environments</li><li>Denver and Aurora schools receive $350K grant to pilot Premo AI tool for multilingual math instruction</li></ul>
SMX plans AI platform launch in 2027
SMX (Security Matters) plans to introduce its AI platform internally in Q1 2027 and reach full functionality in Q2 2027. A limited commercial version is also planned for Q2 2027. No financial terms or revenue estimates were provided yet.
SMX builds AI platform for material research
SMX (Security Matters) is developing SMX Scientific Intelligence, an AI-powered platform to speed up material research and industrial applications. The platform will combine AI with scientific data to help identify chemical agents, additives and molecular markers. Internal use is planned for Q1 2027, with full operations and a limited commercial version targeted for Q2 2027. It will also support recycled plastic certification and the Plastic Circularity Credits marketplace.
SMX advances AI platform for material discovery
SMX is advancing its Scientific Intelligence platform, an AI-powered tool for material exploration and scientific screening. Internal introduction is planned for Q1 2027, with full operation expected in Q2 2027, subject to testing. A limited commercial service is also planned for Q2 2027. Certified recycled-plastics standards have not yet been agreed upon or adopted.
AI could remove 70 percent of bank risk tasks
A McKinsey report says AI could eliminate 60 to 70 percent of manual tasks in bank risk functions. Banks could reinvest freed capacity into risk analysis and strategic decisions. Early adopters may gain competitive advantages, but workforce and governance changes will be needed.
Banks increase AI spending despite governance gaps
A survey of 300 risk and treasury leaders across 31 countries found that 59 percent of banks have embedded AI in decision-making. However, only 18 percent have fully operational AI governance in place. Most banks still restrict autonomous AI decisions, but 76 percent expect scaled use by 2031.
Red Hat layers masked keys and sandboxes for AI agent security
Red Hat is building security tools for AI agents that take actions instead of just answering questions. Its OpenShell runtime gives agents masked credentials so they never hold real keys. Agents can also run inside Kata Containers or virtual machines on OpenShift. Red Hat is working with OpenAI Group on standards for agent identity and authorization.
AST SpaceMobile targets $1B revenue and AI partnerships with Google and Meta
AST SpaceMobile expects revenue near $1 billion in the 12 months after commercial service starts. CEO Abel Avellan said the company is working with Google and Meta on satellite-based AI for cars, drones and IoT devices. Q3 revenue was $42.3 million, missing estimates of $46 million. The company plans intermittent service by year-end 2026 and consumer beta trials in early 2027.
Crusoe sells unfinished Wyoming data center to Google for $17B buildout
Crusoe sold an unfinished data center project in southeastern Wyoming to Google. Google plans to invest $17 billion to develop the campus after concerns about cost and construction timelines. The project was originally planned to reach 10 gigawatts of capacity. Crusoe continues to work with Google on another data center near Amarillo, Texas.
US stock market crash risks rise as AI investment faces new challenges
US stock market risks are growing as AI infrastructure investment slows. About 158.6 gigawatts of planned data center capacity sits in states considering development pauses. Texas Governor-elect Gina Hinojosa has called for a moratorium on new data center approvals. The Federal Reserve has noted wider credit spreads for AI-related assets. The S&P 500's Shiller P/E ratio is near its second-highest level in 156 years.
Alterra Labs raises $12.5 million to train medical AI in simulated environments
Alterra Labs raised a $12.5 million seed round led by Oak HC/FT. The New York startup builds simulated medical environments to test and train AI models. Doctors and scientists design tasks and judge model decisions to help developers improve systems. The company also creates environments for medical imaging, clinical trials and telehealth.
2CRSi partners with UET to deploy AI infrastructure
2CRSi announced a technical partnership with Universal Edge Technologies (UET), a U.S. specialist in AI computing systems. The collaboration focuses on deploying immersion-cooled IT infrastructure. 2CRSi validated UET's VectorEdge Large system and will offer warranties of up to three years for integrated servers. The partnership aims to provide turnkey GPU computing infrastructure for industrial, isolated, or military environments without dedicated data centers or external air conditioning. UET will use 2CRSi's operations in the United States and France to target U.S. and allied-country markets, particularly in government and defense sectors.
AI labs shift from data labeling to simulated worlds
AI research labs are moving away from traditional data labeling and toward simulated environments called RL environments. These simulated workplaces allow AI agents to use tools and get scored on results. Handshake CEO Garrett Lord explained that the old work involved college students drawing boxes around cars, while the new work has experts authoring realistic banks, hospitals, and law firms task by task. Handshake launched Handshake AI in January 2025 and has reported more than $500 million paid out. Labs now evaluate models based on whether small open models can climb internal benchmarks after post-training.
Denver and Aurora schools receive $350K AI grant
DSST schools in Denver and Aurora will receive $350,000 from Multilingual by Design, a Washington, D.C. based philanthropic group backed by the Bezos Family Foundation. The grant will fund a pilot of the Premo AI tool, which helps K-8 teachers plan grade-level math lessons for multilingual students. The tool will roll out across DSST's 16 campuses in Denver and Aurora, where more than 65% of students are current or former multilingual learners. Throughline Learning, a Rhode Island-based nonprofit, will partner with DSST to implement the tool and provide technical assistance to teachers.
Sources
- SMX (Security Matters) Public Limited (SMX) Targets Q2 2027 AI Platform Commercial Launch
- SMX Develops AI-Powered Scientific Intelligence Platform to Accelerate Material Research and Expand Commercial Services
- SMX plans an AI service to help companies find materials for products.
- AI could eliminate up to 70% of manual risk tasks in banks, free capacity for strategic roles
- Global banks increase AI investment amid governance gaps
- AI agent security sees Red Hat layer masked keys and sandboxes
- ASTS Stock Retreats On Q3 Sales Miss – But AST SpaceMobile’s $1B Revenue Opportunity, AI Partnerships Keep Growth In Focus
- Crusoe Sells Unfinished Wyoming AI Data Center Project to Google, Which Plans $17B Investment
- US Stock market crash risks rise as AI investment faces new challenges
- Exclusive: Alterra Labs announces $12.5 million seed round to build training environments for medical AI
- 2CRSi expands its presence in the artificial intelligence infrastructure market
- Labs no longer buy labels, they buy worlds
- Denver, Aurora schools to get $350K for AI tool
Comments
Please log in to post a comment.