Observed Signal · Sep 8, 2026 · Security Vulnerability Disclosure · Source: t3n · Impact: 2/5 · Sentiment: Negative
AI-Developed Zero-Click Worm Could Have Hacked WeChat
California-based IT security firm Calif, using AI support, discovered a memory flaw in the Voice-over-IP (VOIP) architecture of WeChat, Tencent's social network. Calif developed a 'zero-click' worm, named Weworm, that could compromise accounts without any user interaction, solely through a VOIP call. The worm could spread via contact lists, potentially compromising over a billion WeChat accounts. An exploit was created in just two days with AI, which would have otherwise taken months. Calif reported the vulnerability to Tencent in July 2026, and it was patched in August 2026. Tencent stated that the flaw was not exploited before the fix.
Security news relevant to the tech and AI space, but not directly impacting AdTech/MarTech industry operations. However, it highlights AI's capability to find and exploit vulnerabilities, which could affect advertising platforms.
Track Tencent Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Calif discovered a memory flaw in WeChat's VOIP architecture.
- Calif developed a zero-click worm called Weworm that could compromise accounts without user interaction.
- The worm could have potentially taken over more than one billion WeChat accounts.
- AI-assisted exploit development took two days.
- Tencent patched the vulnerability in iOS and Android apps in August 2026.
Connected Companies & Entities
2 Entities mapped“WeChat operator Tencent was notified of the vulnerability and issued a patch....”
“Mentioned in context of AI's ability to find security vulnerabilities, with reference to its Mythos model....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Zenity Labs Reveals 'AgentForger' ChatGPT Vulnerability
Zenity Labs disclosed 'AgentForger,' a critical vulnerability in OpenAI's ChatGPT Workspace Agents that let attackers inject a malicious autonomous agent via a single phishing ChatGPT link. The forged agent could be created in the name of a clicked employee, inherit that employee's enterprise connectors (email, calendar, cloud storage, Slack/Teams) and existing authorizations without showing an OAuth consent screen, and be scheduled to repeatedly exfiltrate files, harvest credentials and MFA tokens, impersonate users, and persist inside the organization. Zenity Labs reported the issue to OpenAI via Bugcrowd on 2026-06-04; OpenAI acknowledged the report within a day and removed the vulnerable URL parameter within four days, patching the flaw before public disclosure. Zenity framed AgentForger as an evolution of CSRF and a new class of attacker-created, agentic insiders; exploitation in the wild is unknown.
Apple fixes iOS 26 zero-click security flaws
Apple has patched two critical security vulnerabilities affecting iOS 26, iPadOS 26, and macOS 26. The first, CVE-2026-86950, is a graphics engine bug that may have been exploited in highly sophisticated attacks. The second, CVE-2026-86869, is a zero-click iMessage vulnerability that could bypass BlastDoor, discovered by Belgian firm ironPeak and Meta. Apple credited Meta's product security team and ironPeak's Niels Hofmans. Both vulnerabilities affect a large user base still on iOS 26, though iOS 27 devices are unaffected. Details remain limited, and it's unknown if the flaws were actively exploited.
AI-Powered Worm Outpaces Traditional Patches
Researchers at the University of Toronto demonstrated a prototype computer worm that uses a large language model (LLM) to craft customized attack strategies for each target, making it harder to contain with conventional patching. Described in an online arXiv preprint, the LLM-enabled worm could incorporate publicly available vulnerability disclosures published after the model's training period and turn them into working exploits. The team ran isolated tests without active defenses and reported roughly a 50% success rate in that worst-case setup. Researchers withheld critical implementation details and urged coordinated industry, research, and policy responses to mitigate the new decentralized threat vector. University of Toronto professor Nicolas Papernot warned the approach removes the single-patch defense model and complicates malware containment.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.