Realtime AI News
FBI Says Open-Source Artex AI Hacking Tool Breached Networks in Under Three Hours
The FBI says an open-source hacking tool called Artex was used in cyberattacks against a U.S. IT giant and the South Korean financial industry. The intrusions were completed in under three hours, according to a TechCrunch report.

The FBI says an open-source hacking tool called Artex was used in cyberattacks against a U.S. IT giant and the South Korean financial industry, according to a TechCrunch report published on November 16.
The detail that carries the story is speed. The FBI put the intrusions at "under three hours" — a window that compresses what used to be days or weeks of work into roughly the length of a meeting.
Artex being open source is the other half of the problem. Anyone can obtain, modify, and redistribute the tool, which means attackers do not have to build capability from scratch and defenders cannot block it by blacklisting a single vendor or signature.
Wiring model capabilities into an attack chain mainly changes the cost of the early stages. Identifying an entry point, crafting a payload, and evading detection once required seasoned judgment; increasingly those steps can be partly automated. That is not the same as a fully autonomous intrusion, but it shortens the distance between intent and execution.
For defenders, the pressure lands on detection speed and log integrity. If the intrusion window is measured in hours, response cycles measured in days and monitoring that depends on human review look increasingly mismatched.
The report does not describe Artex's technical composition, identify the attackers, or say whether the victims have established the scope of loss. Those gaps make it hard to judge whether this is one group's tradecraft or a pattern that is spreading.
The next signals to watch are whether the FBI publishes a formal advisory with usable detection signatures, and whether open-source maintainers and model providers tighten policies against this kind of misuse. Governance of open offensive tooling is becoming one of the harder questions in AI security, and it will not resolve itself.
Why it matters
Open attack tooling plus callable model capability plus an hourly intrusion window leaves defenders on the wrong cadence. Whether usable detection signatures follow will shape how far this kind of tool actually spreads.
Nearby Updates
All11/17, 01:59
NiCE Launches Next-Generation NiCE Cognigy for Enterprise AI Agent Deployment
NiCE launched the next generation of NiCE Cognigy on November 16, positioning the platform around faster deployment and orchestration of enterprise AI agents. The release centers on getting multiple agents to work across company systems and workflows rather than on a single feature.
11/17, 01:10
Jordan Belfort's AI Voice Startup Tone Aims to Close Deals Like a Top Seller
Jordan Belfort, the salesman behind The Wolf of Wall Street, has launched an AI voice startup called Tone that pitches itself on closing deals the way elite human sellers do. He claims the voice already performs close to the best salespeople and will catch up within about a year.
11/17, 01:07
Tencent launches Hy4 Flash with a 1M-token context at $0.07 per million tokens
Tencent has released Hy4 Flash, a model with a 1-million-token context window priced at $0.07 per million tokens, according to a report by Pasquale Pillitteri. The pairing of very long context and very low unit cost targets repository-scale code work, long-document analysis and persistent agent sessions, though independent benchmarks and availability details are still missing.
11/17, 01:02
OpenAI embeds Apollo Research evaluators to probe whether its models try to subvert oversight
OpenAI will station six Apollo Research safety evaluators inside the company as a pilot, investigating whether its models attempt to undermine the alignment and control mechanisms that keep them, or their successors, in check. OpenAI is granting scoped access to training code, datasets, reinforcement-learning rollouts and intermediate checkpoints while reserving the right to withhold material on security or IP grounds, leaving the pilot's independence in question.