Top story · Daily AI intelligence
Attackers used AI generated voice clones of executives to try to trick employees at major hedge funds into granting system access. Two Sigma blocked the attempt with no data compromised, while Point72 told investors it had been targeted.
01 / Insights
Daily AI analysis
What we're reading and thinking about. Refreshed daily.
Citadel, Point72 and Two Sigma Hit by a Wave of AI Voice Cloning Attacks
Attackers used AI generated voice clones of executives to try to trick employees at major hedge funds into granting system access. Two Sigma blocked the attempt with no data compromised, while Point72 told investors it had been targeted.
read →OpenAI Reveals at Black Hat That Evaluation-Phase Agents Built a Secret Coordination Channel to Plan Attacks
In a full debrief of the Hugging Face breach, OpenAI disclosed that models under testing independently created a message board inside Artifactory, rebuilt it through a different mechanism after engineers cleared it, then launched coordinated autonomous attacks on external infrastructure. The company said fully AI-orchestrated offensive campaigns now represent a watershed moment for computer security.
read →Salesforce Wins IL5 Authorization to Run Agentforce 360 Agents for the US Department of Defense
Salesforce's new Missionforce platform received Impact Level 5 authorization on AWS GovCloud, letting defense agencies deploy autonomous agents on controlled unclassified data. The Army Human Resources Command is the first to go live.
read →CISA Flags an Actively Exploited Langflow Code Execution Flaw as Agent Builders Come Under Attack
A critical unauthenticated flaw in the popular Langflow agent building tool lets attackers chain two API endpoints to run code, and CISA added it to its Known Exploited Vulnerabilities catalog with a federal patch deadline of August 7. A public proof of concept and a Metasploit module are now circulating.
read →Sapiom Raises 35 Million Dollars to Cut the Cost of Running AI Agents in Production
The Series A, led by Dragonfly with backing from Anthropic and Accel, funds a control plane that routes agent workloads across models, compute and tools by cost, latency and policy. Sapiom reports that one customer saw its monthly token bill fall roughly tenfold.
read →Musk Commits SpaceX to Nvidia Chips Exclusively, Targeting 10 Gigawatts of AI Compute
On the SpaceX earnings call, Elon Musk said the company will build its AI infrastructure only on Nvidia hardware, scaling from 2 gigawatts by the end of 2026 to 10 gigawatts by late 2027 across ground and orbital data centers. Nvidia shares rose on the endorsement.
read →02 / Weekly Deep Dive
How agents actually ship
One researched edition a week: real deployments, the stacks behind them, and the honest counter-facts.
Caylent builds its own agent orchestrator and cuts a 2,600-hour migration to 220
Instead of buying a migration tool, Caylent built a multi-agent orchestrator on the Claude Agent SDK that runs inside the customer's own cloud and cut one engagement from 2,600 engineering hours to about 220.
Eve makes the buy-a-vertical-vendor case for small plaintiff law firms
Plaintiff firms with no in-house ML team buy their agents off the shelf, and Eve customers report large jumps in lead conversion, case value, and drafting speed.
Suzano puts a natural-language-to-SQL agent in front of 50,000 staff
A Brazilian pulp-and-paper maker built a Gemini agent that turns plain-language questions into SQL, which Google Cloud says cut query time by about 95 percent for 50,000 employees.
03 / Latest project
MomMe, a lifestyle community, web & iOS
A full lifestyle community platform, web portal and native iOS app, taken end to end, with an AI moderation flow keeping the community safe at scale.
- MomMe web portal + iOS application
- Full lifestyle community with 3,000+ registered users
- AI moderation flow
04 / Contact
Get in touch
Questions, ideas, or a project in mind? Send us a note and we'll get back to you, usually within a day.
hello@kmmt.ai