Cisco Adds Claude Mythos to AI Defense Platform

Key Takeaways

• Cisco added Anthropic's Claude Mythos to its AI Defense platform for red-teaming AI applications and agents, announced at Cisco Live APJC in Melbourne
• New Splunk tools include Tokenomics for real-time cost tracking and Eval-G1 for quality evaluation of AI agents
• Cisco launched AI Defense in January 2025 and has helped hundreds of customers find flaws in AI systems
• Stifel CEO Ron Kruszewski said integrating Claude AI into advisor workflows led to more job openings at the firm
• Anthropic amended its S-1 filing to offer 825 million shares ahead of its Nasdaq debut
• Congress faces a Dec. 11 deadline to renew CISA 2015, the Federal Cybersecurity Enhancement Act and the Technology Modernization Fund
• A coalition of 35 industry groups wants CISA 2015 extended through 2035
• Toledo AI Week offers free AI training and certification with over 120 speakers across three venues
• Jobs requiring AI skills carry an average 62% wage premium according to PwC's 2026 Global AI Jobs Barometer
• 73% of B2C marketers using AI agents don't measure them on sales, revenue or conversions

Cisco launches AI agent security and cost tools in Splunk

Cisco introduced new tools in its Splunk platform to help organizations check the security and operating costs of AI agents. The tools include an AI agent security testing tool and a cost monitoring tool. The security tool finds vulnerabilities while the cost tool tracks spending in real time. The tools are available now with a free trial for interested organizations.

Cisco adds Claude Mythos to AI Defense for stronger security testing

Cisco announced it is adding Claude Mythos to its AI Defense platform to help red team AI applications and agents. The update lets security teams stress-test AI systems with advanced reasoning. New Splunk Agent Observability and Cisco IQ improvements also came with the announcement. Cisco said these changes help enterprises secure and govern AI agents at machine speed. Jeetu Patel, Cisco's chief product officer, said the goal is to protect agents while ensuring every token spent has value.

Cisco to use Claude Mythos to secure AI agents and operations

Cisco will add Anthropic's Claude Mythos AI model to its security platform. The move aims to counter AI-enabled cyberattacks and improve the security and efficiency of AI agents. The announcement came at Cisco Live APJC in Melbourne, Australia. Claude Mythos will run multi-step security tests on AI applications and agents. Cisco also introduced Tokenomics for cost tracking, Eval-G1 for quality evaluation, and expanded Cisco IQ and Splunk integration. These tools will help businesses manage AI agent security and spending.

Cisco AI Defense adds Claude Mythos for advanced security testing

Cisco announced it is adding Claude Mythos to AI Defense for red teaming AI applications and agents. The platform helps defenders stress-test AI systems with advanced reasoning. New Splunk Agent Observability features include Tokenomics for cost governance and Eval-G1 for quality evaluation. Cisco launched AI Defense in January 2025 and has helped hundreds of customers find flaws in AI systems. The updates aim to give organizations clear visibility into AI agent costs, performance, and security.

Cisco adds Claude Mythos to AI security testing platform

Cisco is adding Anthropic's Claude Mythos model to its AI Defense platform. The update helps enterprises find vulnerabilities in AI applications and agents. It was announced at Cisco Live APJC in Melbourne on November 10. The release also includes improvements to Splunk's agent observability tools. New integrations between Cisco IQ and the Splunk Platform will help security teams spot and fix weaknesses faster.

Lame Duck Congress Faces Deadline on Cybersecurity and AI Safety Bills

Congress returns with expiring cybersecurity authorities and unfinished AI safety bills. The Cybersecurity Information Sharing Act of 2015, or CISA 2015, expires again on Dec. 11. The Federal Cybersecurity Enhancement Act and the Technology Modernization Fund also expire that day. A coalition of 35 industry groups wants CISA 2015 extended through 2035. Lawmakers also weigh AI proposals like the FRONTIER Act and the AI Kill Switch Act before January leadership changes.

Congress Faces Year-End Deadline for Cybersecurity and AI Legislation

Lawmakers return to Washington with a Dec. 11 deadline for the Cybersecurity Information Sharing Act of 2015. This law protects companies that share threat data with the government. Congress also considers renewing the Federal Cybersecurity Enhancement Act and the Technology Modernization Fund. These items are tied to the National Defense Authorization Act. Several AI proposals are also in play, including the FRONTIER Act and the AI Kill Switch Act.

Toledo AI Week Opens With Free AI Career Training as Demand for AI Skills Soars

Toledo AI Week began Nov. 10 with free AI training and certification through AI Ready Ohio. Jobs that require AI skills carry an average 62% wage premium, according to PwC's 2026 Global AI Jobs Barometer. The event runs through Nov. 12 with sessions at Bowling Green State University, downtown Toledo and the University of Toledo. More than 120 speakers join the Great Lakes AI Week. Attendees can earn AI Explorer, AI Empowered Professional and AI Empowered Leader certificates.

Great Lakes AI Week 2026: Free AI Training in Toledo

Great Lakes AI Week kicked off Tuesday in Northwest Ohio with three days of free training. Events are held at Bowling Green State University, One SeaGate in downtown Toledo and the University of Toledo. The program is powered by JobsOhio and the Enterprise Technology Association. Participants can earn AI Ready Ohio certificates in AI Explorer, AI Empowered Professional and AI Empowered Leader. The week includes workshops, coding sessions and an AI for Business Summit.

Stifel CEO: Using Anthropic's AI Tools Led to More Job Openings

Stifel CEO Ron Kruszewski discussed adding Anthropic's Claude AI into the firm's financial advisor workflow. He said using the tool showed the company needs more positions. The comments came as Anthropic amended its S-1 filing to offer 825 million shares ahead of its Nasdaq debut. The discussion appeared on 'The Claman Countdown.'

CoRA-Bench evaluates evidence preservation under token budgets

Researchers introduced CoRA-Bench, a benchmark for evaluating evidence preservation during long-input inference. The study compared various context selection methods including head and tail truncation, BM25, and dense retrieval across multiple datasets. Results showed task-dependent tradeoffs in evidence preservation across different retrieval methods.

Paloren outlines practical buyer checklist for AI training companies in Australia

Paloren released guidance for Australian businesses choosing AI training companies for 2027. The checklist emphasizes practical exercises, source checking, and workplace trials to evaluate training effectiveness. Training formats include half-day and full-day workshops, department programs, and executive coaching across various AI platforms.

Poesis plans to open flagship agentic equity strategy to external investors

Poesis, an AI-native asset manager, plans to open its agentic equity strategy to external investors in early 2027. The strategy, named AI Clarity, uses dual mandate approach to pursue equity upside while seeking downside protection. The firm, founded by former Capital Group and Goldman Sachs executives, has been trading with proprietary capital since August 2026.

73% of B2C marketers using AI agents don't measure them on sales

Research shows most B2C marketers running AI agents focus on engagement metrics rather than sales outcomes. Only 27% measure AI agents on revenue, bookings, or conversions. Three in four cannot track whether agent-handled leads turn into sales. The findings come from the Invoca 2026 Broken Funnel Report surveying 451 US B2C marketing leaders.

Abnormal AI behavioral security platform focuses on AI-driven cyberattacks

Abnormal AI, founded in 2018, positions itself as a behavioral security platform stopping AI-driven cyberattacks. The company, led by Tony Burnside in Asia Pacific, applies lessons from Adtech to detect attacks exploiting human patterns. Focus areas include phishing, business email compromise, and account takeover beyond traditional security measures.

DKK 325m investment backs national AI platform on Gefion

A DKK 325m investment gives the investor a 23.7% stake in DCAI. The funds will expand Gefion's compute infrastructure and capabilities. DCAI plans to build a secure, scalable national AI platform. The 2026 guidance remains unchanged.

Sources

• Cisco Introduces Tools To Secure AI Agents
• Cisco Expands AI Security with Claude Mythos and Splunk Updates
• Cisco to add Anthropic Claude Mythos to harden AI agent security and operations
• Cisco AI Defense Integrates Claude Mythos for Advanced AI Security Testing
• Cisco adds Anthropic's Claude Mythos to AI security testing platform
• Lame Duck Congress Faces Deadline on Cybersecurity and AI Safety Bills
• Congress faces year-end deadline for cybersecurity and artificial intelligence legislation
• Toledo AI Week Opens With Free AI Career Training as Demand for AI Skills Soars
• Great Lakes AI Week 2026: Free AI Training in Toledo
• Stifel CEO: Using Anthropic's AI tools made us realize we need more positions
• CoRA-Bench: Evaluating Evidence Preservation by Context Selectors Under Token Budgets
• Best AI Training Company in Australia for 2027: Paloren Sets Out a Practical Buyer Checklist
• AI-Native Asset Manager Poesis Plans to Open Flagship Agentic Equity Strategy to External Investors
• New Research: 73% of B2C Marketers Using AI Agents Don't Measure Them on Sales
• Abnormal AI -The behavioural security platform company
• NETC: DKK 325m investment backs a secure, scalable national AI platform built on Gefion