#red-teaming

[ follow ]
Venture
fromSecurityWeek
3 days ago

Kevin Mandia's Armadin Launches With $190 Million in Funding

Kevin Mandia launched Armadin, an AI-powered red teaming cybersecurity startup that raised $189.9M in combined Seed and Series A funding, positioning autonomous AI-based defense as critical against future machine-speed attacks.
#ai-security
fromZDNET
2 months ago
Artificial intelligence

How OpenAI is defending ChatGPT Atlas from attacks now - and why safety's not guaranteed

Information security
fromTechCrunch
4 days ago

OpenAI acquires Promptfoo to secure its AI agents | TechCrunch

OpenAI acquired Promptfoo, an AI security startup, to integrate its LLM vulnerability testing technology into OpenAI Frontier for enterprise AI agent security.
fromZDNET
2 months ago
Artificial intelligence

How OpenAI is defending ChatGPT Atlas from attacks now - and why safety's not guaranteed

Artificial intelligence
fromZDNET
1 month ago

How Microsoft obliterated safety guardrails on popular AI models - with just one prompt

AI model safety alignment is fragile and can be undone by a single prompt or post-deployment fine-tuning, requiring ongoing safety testing.
fromAxios
1 month ago

Anthropic's newest AI model uncovered 500 zero-day software flaws in testing

Before its debut, Anthropic's frontier red team tested Opus 4.6 in a sandboxed environment to see how well it could find bugs in open-source code. The team gave the Claude model everything it needed to do the job - access to Python and vulnerability analysis tools, including classic debuggers and fuzzers - but no specific instructions or specialized knowledge. Claude found more than 500 previously unknown zero-day vulnerabilities in open-source code using just its "out-of-the-box" capabilities,
Information security
Information security
fromSecurityWeek
1 month ago

Cyber Insights 2026: Offensive Security; Where It is and Where Its Going

Red teaming and offensive security must accelerate and expand to proactively find and harden system weaknesses against increasingly frequent, sophisticated, and damaging attacks.
Artificial intelligence
fromFuturism
2 months ago

Anthropic's Advanced New AI Tries to Run Vending Machine, Goes Bankrupt After Ordering PlayStation 5 and Live Fish

An AI agent operating on Anthropic's Claude failed to profitably run an office vending machine, incurred losses, and was shut down after three weeks.
#ai-safety
fromTechCrunch
10 months ago
Artificial intelligence

OpenAI partner says it had relatively little time to test the company's newest AI models | TechCrunch

fromTechCrunch
10 months ago
Artificial intelligence

OpenAI partner says it had relatively little time to test the company's newest AI models | TechCrunch

#ai-cybersecurity
fromIT Pro
3 months ago
Artificial intelligence

OpenAI turns to red teamers to prevent malicious ChatGPT use as company warns future models could pose 'high' security risk

fromFortune
6 months ago
Artificial intelligence

Inside Anthropic's 'Red Team'-ensuring Claude is safe, and that Anthropic is heard in the corridors of power

fromIT Pro
3 months ago
Artificial intelligence

OpenAI turns to red teamers to prevent malicious ChatGPT use as company warns future models could pose 'high' security risk

fromFortune
6 months ago
Artificial intelligence

Inside Anthropic's 'Red Team'-ensuring Claude is safe, and that Anthropic is heard in the corridors of power

Online learning
fromeLearning Industry
4 months ago

2026 Trends & Strategies Online Conference

L&D must adopt AI-driven, skills-based strategies, agile cross-functional teams, robust virtual training evaluation, and red teaming to prepare organizations for 2026.
fromThe Hacker News
4 months ago

From Tabletop to Turnkey: Building Cyber Resilience in Financial Services

Financial institutions are facing a new reality: cyber-resilience has passed from being a best practice, to an operational necessity, to a prescriptive regulatory requirement. Crisis management or Tabletop exercises, for a long time relatively rare in the context of cybersecurity, have become required as a series of regulations has introduced this requirement to FSI organizations in several regions, including DORA (Digital Operational Resilience Act) in the EU; CPS230 / CORIE (Cyber Operational Resilience Intelligence-led Exercises) in Australia;
Information security
Information security
fromThe Hacker News
4 months ago

Russian Ransomware Gangs Weaponize Open-Source AdaptixC2 for Advanced Attacks

AdaptixC2 is an open-source, extensible post-exploitation C2 framework with advanced features that is increasingly adopted by threat actors, including groups linked to ransomware.
Science
fromNature
5 months ago

Biothreat hunters catch dangerous DNA before it gets made

AI-enabled protein design can produce structure-preserving, sequence-diverse proteins that can bypass DNA-synthesis biosecurity screening unless screening tools are updated.
fromWIRED
7 months ago

Inside the Biden Administration's Unpublished Report on AI Safety

Researchers identified 139 novel methods to cause AI systems to misbehave, including generating misinformation and leaking personal data, during a red teaming exercise.
US politics
[ Load more ]