VQV Signal

RESEARCH WATCH TECHNICAL

Sakana AI forms Tokyo Security Research Group and reports Salesforce public-sector deal

According to a post on X by @hardmaru, Sakana AI has launched a Security Research Group in Tokyo, co-led by Akihiro Umegai and @hardmaru. The group will treat security as an upstream quality-engineering problem, focus on multi-agent defense and securing AI agents, and publish its research and benchmarks openly. The post also cites a Nikkei report that Sakana is teaming up with Salesforce to deploy AI models in the public sector.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-11-20T05:30:00+00:00 Detected 2026-11-20T19:00:15+00:00
View original source

According to a post on X by @hardmaru, Sakana AI has launched a Security Research Group in Tokyo, co-led by Akihiro Umegai and @hardmaru. The group will treat security as an upstream quality-engineering problem, focus on multi-agent defense and securing AI agents, and publish its research and benchmarks openly. The post also cites a Nikkei report that Sakana is teaming up with Salesforce to deploy AI models in the public sector.

AI-assisted summary based on the listed source.

Sakana AI's new Security Research Group in Tokyo, co-led by Akihiro Umegai and @hardmaru, will treat security as an upstream quality-engineering problem, research collective multi-agent defense and securing AI agents, and publish its research and benchmarks openly. Separately, Nikkei reports Sakana is teaming up...

Teams building or securing AI agents may find the group's plan to publish research and benchmarks openly useful as a reference, though the post shares no results yet. Public-sector users could be affected if the reported Salesforce deployment goes ahead, but the snippet gives no scope or timing. Watch for the group's first publications and any official confirmation of the Salesforce partnership.

Signal Strength 95% Technical label WATCH Public Interest 40 Category RESEARCH Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 0 Practical Impact Score 36 Novelty Interest Score 94 Consequence Score 50 Curiosity Score 16 Shareability Score 51

VQV surfaced this signal because it is recent, relevant to AI Agents, connected to Scobleizer News Pulse (X).