Fleuret AI, a Paris-based startup, has raised €4 million in a pre-seed round led by RAISE Ventures. The company builds a security platform powered by two AI agents named Emile and Champollion that map environments, find vulnerabilities, and prove they can be exploited. The funding will go toward hiring experts in AI, software engineering, and offensive security. Co-founders Yanis Grigy and Augustin Ponsin want to make security testing a regular part of business operations rather than a one-time check.
Security leaders increasingly say AI is making it harder for companies to keep up with vulnerabilities. Attackers now use AI to find and exploit weaknesses much faster than before, and the pace of attacks has increased according to experts like Rich Mogull of the Cloud Security Alliance. Chief information security officers are turning to offensive security methods such as penetration testing and red teaming to find and fix problems before they are exploited.
Samsung Electronics reported a third-quarter preliminary operating profit of roughly $80 billion, marking the first time the company has surpassed 100 trillion won in quarterly profit. Revenue reached about 195 trillion won, up nearly 127% from a year earlier. Strong demand for memory chips used in AI infrastructure, including DRAM, NAND, and high-bandwidth memory, has driven this growth as prices have soared due to tight supply.
SpaceX has enabled its Grok AI bot to route user questions to multiple external models. Available options include Anthropic's Claude Opus 5.5, MidJourney for images, and Suno for audio. Simple questions route to smaller, faster models while complex ones go to larger models. The move suggests SpaceX may focus more on renting AI infrastructure rather than competing with frontier models, having already invested heavily in building out that infrastructure.
Stanford University launched a new security framework for autonomous AI systems used in teaching, research, healthcare, and administration. The framework is being tested in departments including the Stanford AI Lab and aims to ensure AI agents are transparent, explainable, and accountable. The university plans to expand the framework more widely in the coming months.
Clockwork.io raised $31 million in new funding, bringing its total raised to $73 million. The company builds fault-tolerance software for AI workloads, helping distributed AI training continue when network connections or hardware fail. Its products include LinkPass, which reroutes traffic around failed connections, and TorchPass, which moves work to replacement resources. The company has deployed its technology at LinkedIn and Together AI.
HSBC plans to cut about half of management and specialist roles in its UK wealth business, with reductions among financial advisers reaching around 70 percent. The bank is currently in a consultation period on the proposed changes. CEO Georges Elhedery has made AI a central part of his strategy since taking over in 2024.
Playlab launched its first cohort in North Carolina in late September, partnering with school districts across the state. Leaders gathered to design custom AI tools for schools, learning how to determine if AI fits a task, write system prompts, and test large language models like Gemini, ChatGPT, and Claude. The program aims to build AI agency so students and educators can design and build AI tools rather than just use them.
Elsevier signed a strategic content agreement with the JAMA Network for ClinicalKey AI. Full-text content from 13 JAMA journals will be included, giving clinicians full-text access to more than half of the world's top 100 medical journals through the platform. Cerillion was recognized in the 2026 Gartner Magic Quadrant for CSP AI-Enabled Marketing and Sales Solutions, featuring an AI Management Centre and Agent2Agent support for collaboration between different AI agents.
Key Takeaways
- Fleuret AI raised €4 million for its agentic security platform using AI agents Emile and Champollion to find and exploit vulnerabilities continuously.
- Samsung reported a record third-quarter operating profit of roughly $80 billion, driven by surging AI chip demand for DRAM, NAND, and HBM.
- SpaceX's Grok bot now routes questions to external models including Anthropic's Claude, MidJourney, and Suno, signaling a pivot toward AI infrastructure rental.
- Clockwork.io raised $31 million for fault-tolerance software that keeps distributed AI training running during hardware or network failures.
- Stanford launched an AI security framework for autonomous agents in teaching, research, healthcare, and administration, focusing on transparency and accountability.
- HSBC plans to cut about half of UK wealth management roles and up to 70% of financial adviser positions as part of its AI strategy.
- Playlab's first North Carolina cohort teaches educators to build AI tools using models like Gemini, ChatGPT, and Claude rather than just using them.
- Elsevier partnered with JAMA Network on ClinicalKey AI, offering clinicians access to more than half of the world's top 100 medical journals.
- Cerillion was named in the 2026 Gartner Magic Quadrant for AI marketing solutions with multi-agent collaboration features.
- Security experts say AI has accelerated the pace of cyber attacks, pushing companies toward continuous offensive security practices.
Fleuret AI raises €4 million for continuous security testing
Fleuret AI, a Paris-based company, announced a €4 million pre-seed funding round on October 5. The round was led by RAISE Ventures with support from Auriga Cyber Ventures, Wind Capital, Better Angle, and cybersecurity business angels. The money will help the company hire more staff and develop its platform. Co-founders Yanis Grigy and Augustin Ponsin aim to make security testing a regular part of business operations rather than a one-time check.
AI makes offensive security a continuous business need
Security leaders say AI is making it harder for companies to keep up with vulnerabilities. Attackers can now use AI to find and exploit weaknesses much faster than before. Experts like Rich Mogull of the Cloud Security Alliance say the pace of attacks has increased. CISOs are turning to offensive security methods such as penetration testing and red teaming to find and fix problems before they are exploited.
Fleuret AI secures €4 million for agentic security platform
Fleuret AI received €4 million in funding led by RAISE Ventures. Auriga Cyber Ventures, Wind Capital, Better Angle, and cybersecurity business angels also joined the round. The company plans to use the money to hire experts in AI, software engineering, and offensive security. Its platform uses two AI agents named Emile and Champollion to map environments, find vulnerabilities, and prove they can be exploited.
Samsung forecasts record third-quarter profit of $80 billion
Samsung Electronics reported a third-quarter preliminary operating profit of 107.40 trillion won, which is about $80.2 billion. This is the first time the company has surpassed 100 trillion won in quarterly profit. Revenue reached about 195 trillion won, up nearly 127% from a year earlier. Strong demand for memory chips used in AI infrastructure has driven this growth.
Samsung Q3 profit jumps nearly nine-fold on AI chip demand
Samsung Electronics estimated an operating profit of 107.4 trillion won, about $80.17 billion, for July through September. This marks the company's fourth straight quarter of record operating profit. Revenue is expected to rise 127% to 195 trillion won from a year earlier. Prices for DRAM, NAND, and high-bandwidth memory chips have soared due to tight supply and growing AI demand.
Stanford launches AI security framework for autonomous agents
Stanford University created a new security framework for autonomous AI systems used in teaching, research, healthcare and administration. The framework is being tested in departments including the Stanford AI Lab. It aims to ensure AI agents are transparent, explainable and accountable. The university plans to expand the framework more widely in the coming months.
HSBC plans major UK job cuts driven by AI strategy
HSBC plans to cut about half of management and specialist roles in its UK wealth business. Reductions among financial advisers could reach around 70 percent. The bank is currently in a consultation period on the proposed changes. CEO Georges Elhedery has made AI a central part of his strategy since taking over in 2024.
Cerillion named in 2026 Gartner Magic Quadrant for AI marketing solutions
Cerillion was recognized in the 2026 Gartner Magic Quadrant for CSP AI-Enabled Marketing and Sales Solutions. The company's AI architecture includes an AI Management Centre for governing AI model usage. It also features an MCP Server for accessing BSS/OSS capabilities and Agent2Agent support for collaboration between different AI agents.
Three AI automation stocks show strong revenue growth
AeroVironment focuses on robotic defense technology with its Autonomous Systems segment generating about US$1.4 billion. Ouster supplies lidar sensors for autonomous vehicles and smart infrastructure, generating about US$204.9 million. Teradyne connects AI chip testing with factory automation, with Semiconductor Test bringing in about US$3.7 billion and Robotics around US$356 million.
Elsevier and JAMA Network partner on ClinicalKey AI
Elsevier signed a strategic content agreement with the JAMA Network for ClinicalKey AI. Full-text content from 13 JAMA journals including JAMA, JAMA Network Open and 11 specialty journals will be included. Clinicians will have full-text access to more than half of the world's top 100 medical journals through the platform.
Clockwork.io raises $31 million to protect AI workloads from hardware failures
Clockwork.io announced $31 million in new funding on October 5 to support its fault-tolerance software for AI workloads. The funding was co-led by Premji Invest, Wing Venture Capital, and Seligman Ventures, bringing total funding to $73 million. The company's software helps distributed AI training continue when network connections or hardware fail. Its products include LinkPass, which reroutes traffic around failed connections, and TorchPass, which can move work to replacement resources. The company has deployed its technology at LinkedIn and Together AI.
Grok bot becomes a model router as SpaceX pivots toward AI infrastructure
SpaceX has enabled its Grok AI bot to route user questions to multiple external AI models. Available models include Anthropic's Claude Opus 5.5, MidJourney for images, and Suno for audio. Simple questions route to smaller, faster models while complex questions go to larger ones. The move suggests SpaceX may focus more on renting AI infrastructure rather than competing with frontier models. SpaceX has already invested heavily in building out its AI infrastructure.
Playlab launches first cohort in North Carolina to help schools build AI tools
Playlab launched its first cohort in late September in partnership with North Carolina school districts. Leaders from districts across the state gathered at the Department of Public Instruction in Raleigh to design custom AI tools for their schools. Participants learned how to determine if AI fits a task, write system prompts, and test large language models like Gemini, ChatGPT, and Claude. The program aims to build AI agency so students and educators can design and build AI tools rather than just use them. Playlab says AI readiness is shaping up to be the next digital divide.
Sources
- Fleuret AI Raises €4 Million to Bring Continuous Security Testing to European Businesses
- AI is turning offensive security into a continuous necessity
- Fleuret AI bags €4m for agentic AI offensive security
- Samsung forecasts record third-quarter profit of $80 billion on the back of AI boom
- Samsung Q3 profit jumps nearly nine-fold as AI memory boom lifts chip earnings
- Stanford sets new security rules for autonomous AI agents | ETIH EdTech News
- HSBC plans job cuts across UK wealth business in AI push, FT reports
- Cerillion Recognised in the 2026 Gartner® Magic Quadrant™ for CSP AI-Enabled Marketing and Sales Solutions
- 3 AI Automation Stocks With Revenue Growth Up To 26%
- Elsevier and JAMA Network sign strategic content agreement for ClinicalKey AI
- Clockwork.io Raises $31 Million to Keep AI Workloads Running When Hardware Fails
- Grok Bot Is Now a Model Router as SpaceX Continues to Pivot Further Toward AI Infrastructure
- Playlab launches inaugural cohort in NC, partnering with school districts to help educators and students design custom AI tools
Comments
Please log in to post a comment.