Cisco is rolling out new network tools designed to keep AI agents running smoothly and securely. At its Live Melbourne event, the company introduced features like ThousandEyes intelligence built into Wi-Fi 6E and Wi-Fi 7 access points, along with Experience Metrics for apps and Agentic Actions in Cloud Control. Post-quantum cryptography for wireless networks and expanded Live Protect for switches and routers round out the push to keep infrastructure safe as AI agents take on more responsibilities.
Security remains a central theme. Cisco partnered with Anthropic's Claude Mythos for Adaptive Red Teaming, a tool that stress-tests agent behavior for vulnerabilities. Cisco IQ now monitors close to 20 million devices across more than 16,000 customers. Data Fabric and Splink link live threats to at-risk devices, while new connectors share findings with ServiceNow. An AI-powered playbook also helps teams prioritize infrastructure updates.
In the financial sector, AI agents are reshaping how banks serve customers. A recent analysis argues that institutions relying on personal relationships alone will struggle as AI handles customer queries, recommendations, and lending decisions. Banks must improve operations, adopt cloud and blockchain tools, and adapt risk management. Anthropic's Personal Finance beta gives Claude read-only access to bank accounts through Plaid, though responses vary — Chase blocks external AI agents, while Chime and SoFi welcome them. Wells Fargo now places liability on customers for AI-driven transactions.
Beyond banking and tech, AI adoption is sparking debate elsewhere. HTEC research shows semiconductor companies lead in AI use for manufacturing and design but lack standard ways to measure return on investment. Game designer Masahiro Sakurai said generative AI has not earned his trust, though he remains open to exploring it carefully. A study found nearly 1 in 5 US teens seek mental health advice from AI chatbots, prompting calls for parental involvement and open conversations.
Public pushback is also growing. Voters in 46 of 47 referenda across 13 states rejected data center construction, reflecting concern over hyperscale AI expansion. Some AI firms have taken on massive debt, including $100 billion by August, to build computing infrastructure. Japan's Prime Minister called for global cooperation on AI security during a French forum, while 0sec Labs released an open-source AI security tool called 0 that works across macOS and Linux without a GPU.
Key Takeaways
- <ul><li>Cisco introduced new Agentic AI tools including ThousandEyes, Experience Metrics, and Agentic Actions at Live Melbourne.</li><li>Cisco partnered with Anthropic's Claude Mythos for Adaptive Red Teaming to test AI agent security.</li><li>Cisco IQ monitors nearly 20 million devices for over 16,000 customers with new threat connectors.</li><li>Banks face pressure from AI agents; Chase blocks external access while Chime and SoFi embrace it.</li><li>Anthropic's Claude gains read-only bank access via Plaid, but Wells Fargo shifts liability to customers.</li><li>HTEC finds semiconductor firms lead in AI adoption but lack standard ROI measurement methods.</li><li>Masahiro Sakurai says generative AI has not earned his trust but remains open to cautious exploration.</li><li>Nearly 1 in 5 US teens seek mental health advice from AI chatbots, per recent study findings.</li><li>Voters rejected data center construction in 46 of 47 referenda amid AI infrastructure concerns.</li><li>0sec Labs open-sourced 0, an AI security tool supporting 14 model providers on macOS and Linux.</li></ul>
Cisco unveils stronger network tools for Agentic AI at Live Melbourne
Cisco announced new network capabilities for Agentic AI at Partner Summit. Features include ThousandEyes intelligence built into Wi-Fi 6E and Wi-Fi 7 access points, Experience Metrics for apps, and Agentic Actions in Cloud Control. The company also introduced post-quantum cryptography for wireless networks and expanded Live Protect for switches and routers. These tools aim to make networks faster, safer, and easier to manage as AI agents take on more tasks.
Cisco shows how to keep control as AI agents work faster
Cisco says businesses can let AI agents work fast without losing control by using the network as a central safety system. New tools include Adaptive Red Teaming with Anthropic's Claude Mythos to test agent security. Cisco IQ now monitors nearly 20 million devices for more than 16,000 customers. Cisco Data Fabric and Splunk connect live threats to at-risk devices. New connectors share findings with ServiceNow, and an AI-powered playbook helps teams decide what infrastructure to update first.
Future banks will compete on efficiency rather than customer relationships
A financial industry analysis says banks that rely on personal relationships will fall behind as AI agents reshape financial services. AI agents are already handling customer questions, giving recommendations, and supporting lending decisions. The article argues that banks must improve operations, adopt technologies like blockchain and cloud computing, and become more flexible in risk management. The future of banking is about using technology to improve services rather than replacing human contact entirely.
HTEC says semiconductor firms lead in AI but struggle to measure returns
HTEC research found that semiconductor companies are advanced in adopting AI for manufacturing, product design, and supply chains. However, the industry still lacks a standard way to measure AI return on investment. This makes it hard to connect engineering improvements to real business results. HTEC is calling for a more complete approach to tracking AI impact so companies can make smarter decisions about AI spending.
Kirby designer Masahiro Sakurai says generative AI has not earned his trust
Masahiro Sakurai, designer of Super Smash Bros. and Kirby games, shared his views on generative AI in game development. He said AI has not yet earned his trust and that he prefers creating ideas himself. He warned that AI must not be used to exploit workers. Still, he does not reject AI completely because games are too complex and costly to make. Sakurai is taking a slow approach to learning where and how AI might be useful.
Parents guide teens on safe AI use amid rising chatbot reliance
A study finds nearly 1 in 5 US teens and young adults have sought mental health advice from AI chatbots. Expert Ehrlich says parents should stay involved, explore AI tools with their teens, and keep talking about safe use. She emphasizes that AI should support learning rather than replace a student's own thinking. She also stresses that open family conversations can help people of all ages spot risks like deepfakes.
Voters reject data centers as AI expansion sparks national debate
Voters in 46 of 47 local referenda across 13 states rejected data center construction, reflecting growing concern over hyperscale AI. The article explains that AI firms have taken on massive debt, such as $100 billion by August, to build computing infrastructure. Whether AI succeeds or fails, the author warns it could displace white-collar workers or burden states with debt. The piece calls for federal and state-level action to address the broader economic and environmental risks.
0sec Labs releases open-source AI security tool called 0
0sec Labs open-sourced 0, an AI security harness with a Web UI, CLI, and MCP server. The tool works on macOS and Linux without needing a GPU and lets users bring their own AI model from 14 providers. It is designed to scan web apps, source code, packages, binaries, and kernels for security issues. The harness is free and open source, while model usage costs are paid to providers.
AI may not reach superintelligence, but serious risks remain
The article argues that today's AI systems are unlikely to become superintelligent, but they still pose real dangers. Current models keep refining steps without questioning whether their approach is correct. They also struggle with changing real-world conditions because they learn once and stop updating. The author calls for stronger regulation and independent oversight to reduce the risk of catastrophic outcomes from misuse.
Japan PM urges global cooperation on AI security at French forum
Japan's Prime Minister called for international coordination on AI security during a French conference. The push highlights growing global concern over safe and coordinated AI development. The call comes as governments seek common approaches to managing AI risks. The announcement was reported by Kyodo News on November 10, 2026.
How Banks Can Earn Trust in AI Financial Agents When Big AIs' Read-Only Promise Isn't Enough
Anthropic's Personal Finance beta gives Claude read-only access to bank accounts through Plaid, but banks control the data and consent. Some banks like Chase block external AI agents while Chime and SoFi open their ecosystems. Wells Fargo updated its agreement to place liability on customers for AI agent transactions. Banks should experiment with AI tools and review their own data connections rather than just banning access. Trust is the key asset financial institutions still control directly.
Sources
- Agentic AI demands a stronger, more assured network foundation
- Moving Fast and Staying in Control in the Agentic AI Era
- The Banks of Tomorrow Will Win on Efficiency, Not Relationships
- AFP.com
- Nintendo's legendary Kirby designer Masahiro Sakurai weighs in on gen-AI, saying "it has yet to gain my trust" but not to reject it completely
- How parents can help teens use AI safely & thoughtfully
- An Effective AI Pushback Has to Go Beyond the Local
- 0sec Labs Open-Sources 0, an AI Security Harness With Web UI, CLI and MCP
- Today’s AI won’t achieve superintelligence. But catastrophic risk is real
- BREAKING NEWS: Japan PM calls for int'l coordination on AI security in French confab
- How Banks Can Earn Trust in AI Financial Agents When Big AIs' Read-Only Promise Isn't Enough
Comments
Please log in to post a comment.