Fable 5 is back.....run these prompts before July 12th
NetworkChuck·179K views · 1 month ago
Fable 5 — the most powerful AI on earth — got pulled offline by the U.S. government three days after it launched, and it just came back. But for most of us the free window is only a handful of days (through July 7). So I sat down with Daniel Miessler, who's already living inside that window, to figure out the one thing that actually matters: what's worth asking it before it's gone.
This isn't about doing bigger tasks. It's about pointing maximum intelligence at your deepest systems — your AI harness, the security of everything you've shipped, and the single hardest question about what you're actually building toward — so the payoff outlasts the window. Think of it like a super-intelligent alien that can only help for a week: you don't send it on errands, you have it rebuild your roads.
In this video, you'll learn Daniel Miessler's exact list of meta-prompts to run on Fable 5: how to improve the harness that governs all your AI, how to audit the full attack surface of everything you've deployed, and how to ask the model to tell you what you're really shooting at. Whether you're deep into AI already or just trying not to waste the most capable model on earth, this is the prompt playbook for the week Fable is back.
The FBI has seized hundreds of domains associated with NetNut, a residential proxy service operated by publicly-traded Israeli firm Alarum Technologies, following evidence linking the platform to the Popa botnet. The action was coordinated with industry partners and comes weeks after security researchers connected NetNut to Popa, which comprises at least two million compromised devices operated largely without victim consent.
We’re back, melting - we’ve tried shouting, screaming, and throwing things at the Sun, and it is just not working. Before we begin our analysis, we want to be clear - given the number of vulnerabilities fixed (and some not mentioned..), we’ve struggled to have confidence
Favicons present an overlooked attack surface for reconnaissance, as their static nature and minimal changes make them reliable fingerprints for identifying software deployments across the internet. Bishop Fox has developed an AI-assisted pipeline to hash and enrich favicons at scale, demonstrating a practical methodology for leveraging this technique in threat intelligence operations. The research includes insights into the methodology, resulting dataset, and implications for security professionals conducting asset discovery and infrastructure analysis.
A security researcher has publicly released over 30 proof-of-concept exploits through a platform called "Exploitarium" without following responsible disclosure practices. In an interview with Infosecurity, the researcher explained their rationale for bypassing the typical vulnerability disclosure process before releasing exploit code.
Bitdefender researchers have identified a ransomware campaign where cybercriminals impersonate Interpol in phishing emails to deceive victims into opening malicious attachments. The campaign has targeted businesses globally, leveraging the authority and trust associated with the international law enforcement organization to increase the likelihood of successful infection.
GPT-5.5-Cyber autonomously built a sophisticated fuzzing lab for zlib in a single day—work that would typically take security researchers weeks—discovering multiple vulnerabilities through dynamic testing rather than static analysis. Trail of Bits' "Patch the Planet" initiative is racing to find and patch bugs in open-source projects before advanced AI models democratize vulnerability discovery for attackers. The critical takeaway: the technical barrier protecting unmaintained codebases has collapsed, making proactive AI-driven security audits an urgent necessity for all security-critical software.
Apple's June 2026 security update addressed 37 CVEs primarily affecting WebKit and WebRTC components, with the majority being denial-of-service bugs triggered through malicious web content. Two kernel vulnerabilities stand out as critical: CVE-2026-43724 enables kernel memory writes for privilege escalation, while CVE-2026-39868 causes kernel memory corruption and is credited to elite offensive security researchers, suggesting weaponizable potential. A pair of WebKit sandbox-escape vulnerabilities (CVE-2026-43725 and CVE-2026-43701) are particularly concerning as they bridge remote web exploits to potential kernel-level compromise.
The Veil#Drop campaign leverages Google Blogspot as an infrastructure component to deliver the PureLog Stealer infostealer directly into memory, bypassing traditional file-based detection methods. This fileless approach reduces forensic artifacts and allows the malware to operate without writing to disk, making it more difficult for endpoint security tools to identify and remediate.
A Brazilian banking trojan called Ousaban is actively targeting users in Spain and Portugal through phishing campaigns, according to FortiGuard researchers. The malware's expansion into European markets represents a notable geographic shift for the threat actor beyond its original Brazilian focus.
Aflac Japan has disclosed a data breach exposing policy details along with personal and banking information to regulators. The incident impacts millions of individuals whose sensitive data has been compromised.
Cloudflare has launched Attribution Business Insights, a new dashboard designed to help website owners gain visibility into crawler behavior, activity levels, and potential business value. The tool aims to enable data-driven discussions about crawl resource costs and fair compensation between websites and crawlers.
Well, well, well - once again, the cat has dragged us in and spat us out. Today, we find ourselves questioning the reality we sit within. Must it be so predictable, and why us? “But watchTowr, what do you mean?” Well, if you’re here, you likely fit
Nissan has disclosed a data breach affecting employee information that was exploited through a zero-day vulnerability in Oracle PeopleSoft. The breach is linked to an ongoing campaign targeting the Oracle vulnerability to steal sensitive data from organizations.