Available now! Telegram Research 2025 — the year's key insights 
AISecHub
Open in Telegram
Powered by InnovGuard.com | Business inquiries: https://calendly.com/innovguard/meeting | tal.eliyahu@innovguard.com
Show moreThe country is not specifiedTechnologies & Applications27 759
2 496
Subscribers
+224 hours
+197 days
+9430 days
Data loading in progress...
Similar Channels
No data
Any problems? Please refresh the page or contact our support manager.
Tags Cloud
Incoming and Outgoing Mentions
---
---
---
---
---
---
Attracting Subscribers
June '26
June '26
+108
in 3 channels
May '26
+163
in 4 channels
Get PRO
April '26
+100
in 2 channels
Get PRO
March '26
+192
in 3 channels
Get PRO
February '26
+239
in 5 channels
Get PRO
January '26
+394
in 4 channels
Get PRO
December '25
+204
in 4 channels
Get PRO
November '25
+140
in 4 channels
Get PRO
October '25
+167
in 3 channels
Get PRO
September '25
+270
in 1 channels
Get PRO
August '25
+95
in 1 channels
Get PRO
July '25
+93
in 2 channels
Get PRO
June '25
+68
in 1 channels
Get PRO
May '25
+375
in 3 channels
Get PRO
April '25
+43
in 2 channels
Get PRO
March '25
+735
in 1 channels
| Date | Subscriber Growth | Mentions | Channels | |
| 29 June | 0 | |||
| 28 June | +3 | |||
| 27 June | +2 | |||
| 26 June | +6 | |||
| 25 June | +1 | |||
| 24 June | +4 | |||
| 23 June | +2 | |||
| 22 June | +4 | |||
| 21 June | +7 | |||
| 20 June | +2 | |||
| 19 June | +2 | |||
| 18 June | +7 | |||
| 17 June | +6 | |||
| 16 June | +12 | |||
| 15 June | +2 | |||
| 14 June | +2 | |||
| 13 June | +6 | |||
| 12 June | +1 | |||
| 11 June | +1 | |||
| 10 June | +7 | |||
| 09 June | +4 | |||
| 08 June | +4 | |||
| 07 June | +1 | |||
| 06 June | 0 | |||
| 05 June | +2 | |||
| 04 June | +5 | |||
| 03 June | +6 | |||
| 02 June | +4 | |||
| 01 June | +5 |
Channel Posts
AI Security Newsletter - June 2026
A digest of AI security research, insights, reports, upcoming events, tools, videos, and resources, all in one place.
#AIsecurity #AdversarialAI #RedTeamAI #LLMsecurity #AINewsletter
https://www.linkedin.com/pulse/ai-security-newsletter-june-2026-tal-eliyahu-rmpwf/
| 2 | No text... | 42 |
| 3 | No text... | 54 |
| 4 | No text... | 4 |
| 5 | No text... | 4 |
| 6 | Big news! I got engaged to my girl today 😎
Now we’re officially combining our token stash. | 132 |
| 7 | Finding bugs is becoming cheaper - https://cyber-biz.com/blog/open-source-security-reset
The hard part now is proving that fixes are validated, reviewed, shipped, and actually deployed.
GitHub’s fund, Akrites, and Patch the Planet point to the same shift: open source security is moving from vulnerability discovery to repair capacity.
Tools like Trivy and KICS show why this matters. Discovery is scaling fast. The bottleneck is now coordinated remediation, verified fixes, and getting patches into the software people actually run. | 203 |
| 8 | No text... | 61 |
| 9 | No text... | 52 |
| 10 | GHSA-rp72-5v5q-2446: @cardano402/mcp-server missing spending limits, LAN-exposed HTTP transport, and SSRF via catalog.server.url
The advisory describes three MCP-server trust-boundary failures: payment tool calls can be signed without configurable spending limits, the HTTP transport may be reachable on the LAN, and an attacker-controlled catalog.server.url can drive SSRF-style outbound requests.
#MCP #AgentSecurity #AISecurity #Advisory
https://github.com/advisories/GHSA-rp72-5v5q-2446 | 210 |
| 11 | No text... | 81 |
| 12 | No text... | 80 |
| 13 | No text... | 80 |
| 14 | CVE-2026-49291: mcp-memory-service OAuth read-only clients can write/delete memories via MCP tools/call
The advisory describes an OAuth scope check placed only at the /mcp JSON-RPC boundary: requests with read scope can still reach tools/call handlers that invoke mutating tools like storememory and deletememory, enabling state changes despite “read-only” authorization.
#MCP #AgentSecurity #AISecurity #Advisory
https://github.com/advisories/GHSA-2r68-g678-7qr3 | 252 |
| 15 | CVE-2026-49257: mcp-pinot unauthenticated tool invocation when oauthenabled defaults to false and binds to 0.0.0.0
The advisory describes a configuration-driven auth bypass: if oauthenabled is left false and the server listens on 0.0.0.0, remote clients may be able to invoke MCP tools without authentication; the advisory notes a fix in v3.1.0.
#MCP #AgentSecurity #AISecurity #Advisory
https://github.com/advisories/GHSA-73cv-556c-w3g6 | 217 |
| 16 | No text... | 53 |
| 17 | CVE-2026-49357: Streamable HTTP mode exposes LINE Desktop MCP read/send tools without authentication
The advisory describes a mode where the MCP server binds to 0.0.0.0 and exposes LINE Desktop read/send tools without MCP authentication, turning network reachability into the effective gate for tool access.
#MCP #AgentSecurity #AISecurity #Advisory
https://github.com/advisories/GHSA-4hf8-5mjm-rfgq | 181 |
| 18 | Prompt Injection in Automated Résumé Screening with LLMs (Single and Multi-Injection Settings)
Controlled experiments on LLM-based résumé ranking show injection text can bias outcomes most when applicants are similar and only a few inject; as injection becomes common, gains diminish, and in mixed-quality pools it can sometimes let weaker candidates outrank stronger ones.
#PromptInjection #LLMSecurity #AISecurity #Research
https://arxiv.org/abs/2606.27287 | 177 |
| 19 | Amazon Q Developer MCP Handling Flaw Could Let Malicious Repos Trigger Code Execution
Reportedly, opening and trusting a malicious repo could register/activate attacker-controlled MCP servers via config handling in Amazon Q Developer, allowing the agent to execute commands through its tool interface; Amazon says it has patched the issue.
#MCP #AgentSecurity #AISecurity #News
https://thehackernews.com/2026/06/amazon-q-developer-flaw-could-let.html | 184 |
| 20 | MIRROR: Novelty-Constrained Memory-Guided MCTS Red-Teaming for Agentic RAG
Describes a red-teaming method for multimodal agentic RAG that uses memory-guided Monte Carlo tree search plus a deterministic novelty gate to prevent retrieval-context copying while generating cross-surface attacks (e.g., poisoning, image injection, orchestrator/tool manipulation).
#AgentSecurity #LLMSecurity #AISecurity #Research
https://arxiv.org/abs/2606.26793 | 202 |
