Palo Alto Networks, Nearest Neighbor, Microsoft, AI safety taskforce & summit.

[Available Now!] Book Report: Q3 2024 - GenAI Safety and Security

Highlights:

• Palo Alto Networks: 2000 firewalls compromised.

• "Nearest neighbor" WiFi attack revealed.

• Microsoft enhances Windows security.

• New AI safety taskforce formed.

• International AI safety summit held.

Deep Dive

1. 2,000 Palo Alto Networks Firewalls Compromised Help Net Security

• Zero-day vulnerabilities for remote code execution.

• Compromised devices primarily in US and India.

• Attackers deployed webshells, exfiltrated data, & deployed cryptominers.

• Palo Alto Networks Panorama and WildFire appliances also affected.

• Arctic Wolf observed intrusions across various industries.

Mitigation: Apply security patches and follow remediation guidance.

2. "Nearest Neighbor" WiFi Attack Tom's Hardware

• Russian hackers accessed a US firm's WiFi outside a physical range.

• First heard for two different networks to compromised the third one.

• Attackers leveraged both wired and wireless connections.

Mitigation: Limiting Wireless access point, hiding SSIDs, and mandate MFAs.

3. Microsoft Security Enhancements Cybersecurity Dive

• Windows Resiliency Initiative allows for off-line Windows Update changes.

• Safer deployment practices with endpoint security partners are being implemented.

• Moving away from C++ to Rust for safer programming.

• New capabilities to create security products outside kernel mode.

• Available to the Windows Insider Program community starting in early 2025.

• Initiative follows July Crowdstrike global outage affecting 8.5M devices.

Notice: Stay informed and proactively adopt MS updates.

Online spyware, Zero Trust, Pirated browser attack, Singtel Quantum safe, UK AI safety.

4. New AI Safety Taskforce (TRAINS) NIST

• TRAINS focuses on national security implications of AI.

• Experts from Commerce, Defense, Energy, Homeland Security.

• Will coordinate AI model research and testing.

• Aims to maintain US leadership in safe AI development.

• Operationalizes whole-of-government approach to AI safety.

Mitigation: Monitor progress and recommendations for best practices.

5. International AI Safety Summit TIME

• Summit brought together experts from 9 nations and the EU.

• Focus on synthetic content, testing models, and assessments.

• US and UK AISIs shared findings on model vulnerabilities.

• $11 million in funding announced for synthetic content risk mitigation.

• Emphasizes the need for integrating safety with AI innovation.

Notice: The next summit will happen in Paris, France.

Oracle AI security, AI safety congress hearing, UN Global Digital Compact, UN AI advisory body, AI safety summit.

---