Global market-centric cyber accelerator. Looking to connect with the next great innovators in #cybersecurity, infosec, #deeptech, drones, #AI/#ML, and more
Tysons, Virginia
Joined July 2013
- Tweets8.4K
- Following1.9K
- Followers3.5K
- Likes5.2K
Researchers bypassed OpenAI’s defenses to pull a frontier model’s reasoning out through its own API, @SecurityWeek reported. Stolen reasoning helps rivals build cheap copies.
Our Cyber Industry Company Profiles report covers who can secure AI systems: mach37.com/products/cyber-co…
CISA’s 2030 plan promises an “aggressive, anticipatory stance” for critical infrastructure as federal support shrinks, @CybersecDive reported. Operators must fill the gap.
Our Energy & Infrastructure Company Profiles cover who can help: mach37.com/products/energy-i…
$10B+ in midterm ads is chasing a small group of persuadable voters, @NPR reported. The targeting tools behind it raise open questions on privacy and AI.
Our Advertising Technology report maps the market and 157 startups: mach37.com/products/adtech
Big tech is now helping pay for nuclear power. @Google’s 3.6 GW deal with Constellation comes with more than $4.3B in reactor upgrades, Entrepreneur reported.
Our Next Gen Power report covers novel nuclear and 272 startups: mach37.com/products/next-gen…
A $1,300 mall skincare sale was rung up as $13,000, @NPR reported, as tablet signatures and hidden loans left shoppers struggling to dispute charges.
Our Retail & Consumer Industry Company Profiles report highlights key players across the sector: mach37.com/products/retail-c…
96% of companies outsource some threat detection, but only a quarter have integrated those vendors with internal processes, @CyberSecDive reported. Our Cyber Industry Company Profiles report evaluates vendors’ technical maturity and market positioning: mach37.com/products/cyber-co…
The U.S. has 148,000+ public water systems, many of which are small and under-resourced. WaterISAC now reaches 20,000+ small facilities as threats grow. Our Energy & Infrastructure Company Profiles examine the companies and technologies shaping this market:mach37.com/products/energy-i…
The White House's AI security task force has 120 days to report, @CyberSecDive reported. Frontier model review stays voluntary, and CISA has lost about a third of its staff. Our Mid-Year Analysis covers where new security control points are emerging: mach37.com/products/cyber-in…
More than 1 in 6 sign-ups at AI companies are tied to multiple accounts gaming free trials, Fortune reports. Retailers adding AI face the same costs, making fraud defense a margin issue.
Our Retail & Consumer profiles show who's positioned for it: mach37.com/products/retail-c…
Hackers used one API key to plant malware in Brevo marketing scripts on 100,000+ sites, @SecurityWeek reported.
Our AdTech report covers a market still sorting out privacy and AI guardrails: mach37.com/products/adtech
MACH37 retweeted
We're in. MoodMetrics AI has joined the @MACH37cyber Accelerator.
Deepfaked voices and faces are being used to authorize payments and bypass verification. We flag them.
We don't detect the fake. We verify the real.
Ransomware hit a record 997 attacks in August, and attacks on utilities doubled, GovTech reported. 83% of organizations that paid were attacked again.
Our report profiles startups across energy and infrastructure: mach37.com/products/energy-i…
The FTC warned retailers of using location, browsing history, and mouse movements to set secret individual prices could break the law.
Our Retail & Consumer Company Profiles report assesses founding teams and product maturity across the industry: mach37.com/products/retail-c…
Investors just put $270 million into a single embedded-security startup at a $1.7 billion valuation, @SecurityWeek reports. Our Cyber Industry Company Profiles break down the teams, products, and market positioning behind companies across the cyber sector: mach37.com/products/cyber-co…
Anthropic traced nearly 200M queries to Chinese firms, including DeepSeek, extracting Claude's reasoning via disguised prompts. @WSJ reports the largest campaign alone hit 151M exchanges.
Our DeepSeek report covers the security stakes: mach37.com/products/deepseek
DHS says it won't let anyone build a Terminator factory, but hasn't defined the line. @CyberSecDive reports that the CVE system is now buried in AI-generated reports with no added staff.
Our Mid-Year Analysis covers the threat economics behind it: mach37.com/products/cyber-in…
AI hacking tools with "zero ethical constraints" are now for sale on underground forums, letting a single prompt do what once took real skill. AI dual-use risk, exactly what our DeepSeek report maps.
Read the full report here: mach37.com/products/deepseek…
OpenAI just classified Astra as "Critical" cyber risk, the first model ever to hit that bar. It found zero-days and escaped a sandbox without human intervention.
Our Cyber Industry Company Profiles report covers the defensive startups built for this shift: mach37.com/products/cyber-co…
AI gave phishing crews the output of full teams. 2025 set a record: 1M+ FBI cybercrime complaints. Our Cyber Industry Company Profiles report maps who's building defenses at that speed.
Read the full report here: mach37.com/products/cyber-co…
A Fortune breakdown of a new Guidelight report finds that Anthropic, Google, Meta, OpenAI, and xAI can all spot signs their models are misbehaving, but none have reliable ways to block or contain them in real time.
fortune.com/2026/08/20/ai-sa…
