Sign Up - Linkedin

Elections in the United States are just around the corner, while Brazil and Israel have recently held some of their most consequential elections. The stakes are so high that several of my friends traveled all the way from Los Angeles to Israel just to cast their votes, returning home the very next day.

There is a lot at stake, and bad actors know it. AI-powered swarms are creating new opportunities to spread misinformation, manipulate public opinion, and convince voters that elections don’t matter or misrepresent candidates and ballot measures. At the same time, automated hacking agents are increasingly targeting our online accounts, devices, and digital systems.

In this edition, we explore how AI swarms and cyberattacks could threaten elections worldwide, what you should watch out for, and the practical steps you can take today to protect your devices, accounts, and personal information.

In today’s issue:

• 📰 AI News and Trends

• How to Stay Safe in the Age of AI-Powered Hacking Swarms

• 🧰 AI Tools of The Day - Sales Automations Tools

• AI Swarms, Cyberattacks, and Elections. Is Democracy Under Threat?

• 🐈‍⬛ GitHub repos of the week - A Subagent skills

📰 AI News and Trends

• Mistral released the weights of Le Chonk, advancing the frontier of open models in agentic coding and cybersecurity

• OpenAI’s updated Model Spec, introduces five core values to guide how AI models behave: Faithfulness to Human Intent, Minimizing Harm, Honesty and Transparency, Human Agency, and Care

• AI can turn security fixes into working exploits before users receive an update

• A new Raindrop study involving 2,160 teams of AI agents found that AI can coordinate without assigned leaders, but only 5% of teams developed a clear leader. Claude Sonnet 5.5 and GPT-6.1 Sol were more likely to lead, while agents listed first in prompts were 3 times more likely to propose plans. Even small changes in how AI agents are configured can dramatically influence their behavior

• AI coordination, not just intelligence, could reshape the global economy. Google DeepMind researchers argue that AI’s biggest economic transformation will come not from smarter individual models, but from AI agents working together effectively. Their research identifies four stages: minimal AGI, supercompetent AI, supercooperative AI, and artificial superintelligence (ASI). As AI coordination improves, companies could scale dramatically with fewer employees, disrupting jobs, wages, and entire industries

• Crypto giant Binance hit by $6.3bn hack suspected to be largest in history

• DoorDash is acquiring grocery-delivery company Instacart for roughly $17 billion

• Some Polymarket user funds stolen by hackers; exchange referring matter to U.K. authorities

AI Swarms, Cyberattacks, and Elections. Is Democracy Under Threat?

With U.S. midterm elections approaching on November 3, a wave of cyberattacks and AI-powered manipulation is raising concerns about election security worldwide.

A critical vulnerability in PHP, the technology powering roughly 70% of websites, has exposed potential security weaknesses across the internet. Meanwhile, researchers warn that AI agents can automate cyberattacks and coordinate misinformation campaigns at unprecedented speed.

The timing is concerning. Brazil and Israel have just held major elections, while United States prepare to vote on Tuesday. The biggest threat may not be hackers changing votes, but convincing voters that election results cannot be trusted.

1. AI swarms could manipulate public opinion

Research highlighted by the Max Planck Institute warns that thousands of coordinated AI agents could impersonate real people, spread political misinformation, and manufacture the appearance of widespread public support.

Unlike traditional bots, these agents could adapt their messages, interact with voters, and operate continuously.

2. Cyberattacks could disrupt election services

Recent PHP vulnerabilities raise concerns about attacks against campaign websites, election information portals, and political organizations.

Hackers could potentially steal sensitive information, publish false announcements, or disrupt access to voting information.

Important: PHP’s widespread use does not mean 70% of websites have been compromised, and a website breach does not automatically compromise voting systems.

3. Critical infrastructure is also at risk

A recent Reuters investigation reported that the Pentagon is investigating suspicious refrigeration failures at U.S. military facilities as possible cyberattacks.

Although no connection to elections has been established, the incident highlights how cyber threats can extend beyond websites into physical infrastructure.

4. The biggest danger: Losing trust in elections

Imagine election-results websites going offline while thousands of AI-generated accounts claim that votes have been manipulated.

Even if ballots remain secure, coordinated misinformation could create confusion, discourage voters, and undermine confidence in the results.

The United States’ decentralized election infrastructure, paper ballots, and auditing procedures provide important safeguards. But defending public trust is becoming increasingly difficult.

Zooming out

AI is changing election interference from a human-intensive operation into something that could potentially be automated at massive scale. The convergence of cyber vulnerabilities, AI agents, and political polarization creates new risks for democracies worldwide.

Can AI convince millions of people that their votes no longer matter? We hope not.

🧰 AI Tools of The Day

Sales Automations Tools

1. Strawberry 1.0 - A browser with role-based AI “Companions,” such as Sales Sally for leads and Marketing Martin for posts. They research prospects in parallel, transcribe meetings and update your CRM or spreadsheets.
   How to use it: Point Sales Sally at a list of 50 target accounts and have it push its findings into a Google Sheet your outreach flow already reads.

2. TheySaid 4.0 - Its synthetic users click through your landing page or app and narrate where they get stuck. You get usability feedback in minutes, without recruiting a panel.
   How to use it: Run a test on every new client landing page before ad spend goes live, and paste the friction report into the revision brief.

3. N8N Chat UI - It turns any n8n AI workflow into a branded website chat widget, with voice, file uploads and streaming. It plugs into HubSpot, Calendly, Shopify and Supabase.
   How to use it: Build an n8n lead-qualifier flow, wrap it in the widget, and embed it on a client’s WordPress or Webflow site so it books calls straight into Calendly.

How to Stay Safe in the Age of AI-Powered Hacking Swarms

The past few days, even after the PHP exploits, I keep seeing attempts to break into my Substack account. These swarm of agents are making thousands of attempts on you Substack account as well as your bank account, and home devices accounts and even your car digital systems.

Over the past few days, even after the recent PHP exploits, I’ve continued to see attempts to break into my Substack account. Swarms of automated AI-powered agents are targeting your online accounts at an alarming scale, making thousands of login attempts on your Substack, banking services, smart home devices, and even connected vehicle systems.

A new wave of AI-powered cyberattacks is raising serious concerns about internet security. Some experts are calling it “Black October,” as hackers use AI agents to discover vulnerabilities, launch automated attacks, and target thousands of websites, banks, and businesses at unprecedented speed and lower cost. Recent reports describe attacks affecting financial institutions, major companies, and critical infrastructure. The biggest concern is that AI is making cybercrime faster, cheaper, and easier to scale, while security teams struggle to keep up. For everyday internet users, the main risks include stolen personal information, compromised accounts, sophisticated scams, and service disruptions.

what can you do:

• Avoid unfamiliar websites, especially for payments and logins.

• Update devices, browsers, and routers using official updates.

• Delete unused apps and disconnect unnecessary smart devices.

• Never share passwords or verification codes.

• Enable 2FA and use unique passwords.

• Beware of scams: Verify bank messages directly.

• Monitor bank accounts and report suspicious activity.

• Keep emergency cash, food, and water.

• Don’t panic: FDIC insurance protects eligible bank deposits.

• Share these tips with family and friends.

🐈‍⬛ GitHub repos of the week