PPhying中文
Curated security research, vulnerabilities, advisories and tools for practitioners.

GPT-4o: Actually Good Multimodal AI

josephthacker.com | blog | #ai-security | #llm | #openai | #multimodal | #gpt-4o | #model-release
📄 Raw .md⬇️ Download .md🔗 Original ↗

Summary

Joseph Thacker comments on OpenAI's GPT-4o, a single multimodal model handling text, images, and audio, calling it genuinely impressive.
Published
Collected

Related coverage

blog · specterops.io
AI for Offensive Security: What Works, What Does Not, and How to Adopt It
SpecterOps maps where AI helps offensive security (recon, tooling, vuln discovery) and where it still fails, stressing that results hinge on the harness: context, tools, state, independent validation, and human judgment.
blog · josephthacker.com
How to Hack AI Agents and Applications
A comprehensive guide to hacking LLM-powered applications covering jailbreaks, prompt injection scenarios, data source and exfiltration mapping, multimodal attacks, and AI bug bounty tips.
blog · blog.trailofbits.com
Introducing Patch the Planet
Patch the Planet, a Trail of Bits-OpenAI initiative, sends expert engineers to fix critical open-source projects alongside frontier models; week one yielded 64 pull requests across 19 projects.
blog · blog.trailofbits.com
Weaponizing image scaling against production AI systems
Trail of Bits shows how image-scaling attacks hide prompt injections visible only after downscaling, enabling data exfiltration on Gemini CLI, and releases the open-source Anamorpher tool.
blog · xbow.com
How Attackers Are Using AI & What Defenders Must Do
Attackers are running AI driven attack chains from recon to exploitation in minutes. Here’s what actually changed, and why defenders need AI pentesting to keep pace.
blog · projectblack.io
Bypassing EDR with Local AI
A pentest shows Claude refuses to build an LSASS dumper, but local DeepSeek and uncensored Qwen models produce a working XOR-encrypted minidumper, making AI-assisted EDR evasion nearly effortless.
Back