🐞Bugs by AI—Caught by AI

👋 Welcome! Today’s breakthroughs show that AI isn’t just exposing risks—it’s unlocking new ways to protect what matters. The future of cybersecurity is being built now, and yes—you’ll want a front-row seat. Let’s unpack the signals shaping tomorrow!

In partnership with

all-in-one platform to manage and scale entrepreneurship communities, opportunities, and innovation.

AI Boom Revives Europe’s Coal Plants – Tech giants like Microsoft and Amazon are turning Europe’s old coal and gas plants into data centers to meet soaring AI energy demands. Utilities including RWE, Engie, and Enel are seizing the chance to lease or operate these sites, benefiting from existing power and water infrastructure while locking in long-term, high-margin power deals. The shift could turn stranded assets into billion-euro AI infrastructure hubs while accelerating renewable energy investments.

AI Bug Hunter Finds 20 Flaws – Google unveiled the first findings from its AI-powered vulnerability researcher, Big Sleep, which discovered 20 security flaws across popular open-source tools like FFmpeg and ImageMagick. Google emphasized that all 20 bugs were found and reproduced by the AI agent alone, highlighting a leap in automated cybersecurity tools. This milestone marks growing trust in AI-led security research—but also raises concerns about hallucinated bug reports.

Orange Boosts AI for Africa – Orange will use OpenAI’s latest models to improve AI tools for African languages, which often lack sufficient data and support. The telecom giant, active in 18 African countries, plans to fine-tune and share these tools for free with local governments to bridge the digital divide.

OpenAI Models Launch on AWS – OpenAI’s latest AI models are now available on Amazon Web Services, letting companies build smarter apps directly on the world’s most-used cloud platform. These open-weight models allow developers to customize and fine-tune the AI for tasks like customer service, research, or complex problem-solving—without starting from scratch. This makes advanced AI more accessible to businesses of all sizes, accelerating the rollout of custom AI tools across industries.

Boeing Fighter Jet Workers Strike – More than 3,200 union members who assemble Boeing’s fighter jets in St. Louis and Illinois walked off Monday after rejecting a contract offering 40 percent average wage growth, a $5,000 ratification bonus, and enhanced vacation and sick-leave provisions.

Cognition Backtracks After Windsurf Deal – Cognition is offering buyouts to nearly all Windsurf employees, just weeks after acquiring the AI coding startup. Though initially praised for securing top talent, Cognition now seems more focused on Windsurf’s technology than its people. Those who stay face intense office demands, highlighting a sharp shift from early promises to a high-pressure, product-first culture.

Nissan European Office Restructuring Talks – Nissan has begun consultations with its European regional office union in France to discuss sweeping restructuring, including potential layoffs among its 560 staff, as part of a broader cost-cutting and efficiency drive across multiple global regions.

Market Turmoil Brings Big Payouts – Wall Street traders are set to receive 10% to 30% higher bonuses this year, fueled by profits from navigating volatile markets amid tariff tensions and economic uncertainty, according to Johnson Associates. While hedge fund executives and asset managers will see more modest gains, investment bankers may face flat pay as deal closings remain slow. With volatility favoring risk-takers, traders have emerged as this year’s clear winners in finance compensation.

AI Giants Cleared for Federal Deals – OpenAI, Google, and Anthropic have been approved to sell AI services to U.S. federal agencies through a streamlined contracting system. Their inclusion in the GSA’s Multiple Award Schedule means faster deployment of AI across civilian departments, with billions in potential government spending now in reach. The decision opens a major revenue stream for AI firms and could accelerate public sector adoption of advanced automation tools.

Small Firms Bet Big on Ether – Small-cap companies are piling into ether, drawn by its staking rewards and role in decentralized finance. Corporate holdings jumped to $3.5 billion by July, as firms see it as a middle ground between bitcoin’s stability and riskier altcoins. But volatility and regulation still cloud its future.

August AI Targets Legal Workflows – Founded by Columbia graduates, AI startup August secured a $7 million seed round led by NEA and Pear VC to automate document-heavy tasks at mid-sized law firms. Backers include OpenAI’s David Azose and Bain Capital Ventures.

Alaan’s Middle East Momentum Builds – Dubai’s Alaan closed a $48 million Series A funding led by Peak XV Partners, marking one of the largest fintech VC rounds in the MENA region. Regional investor interest is shifting toward digital-first financial tools catering to SMEs and startups in MENA.

AI tools now autonomously unearth hidden flaws in software—promising efficiency but also demanding deeper reflection: does delegating this work risk compromising human judgment, autonomy, and governance?

The rise of AI in cybersecurity isn’t just a leap in speed or accuracy—it’s a shift in who (or what) holds responsibility for keeping systems safe. As machine learning models take over tasks once handled by skilled analysts, we’re quietly entering a new phase of digital defense, one where trust, labor, and decision-making are subtly being redefined. The appeal is obvious: tools that scan millions of lines of code, spot vulnerabilities before humans can, and issue alerts with mechanical precision. But that convenience comes with a cost—one we may not comprehend until it's too late.

Recent experiments have shown that advanced AI systems can independently identify serious software vulnerabilities, including those previously unknown to developers. While the discovery rate remains relatively low, the implications are significant—these tools are now capable of performing complex security analysis once reserved for experts. At the same time, growing concerns have emerged about autonomous AI systems that can execute entire cyberattack chains, from initial reconnaissance to exploitation, without human involvement. Adding to the complexity, researchers have warned that these very tools—designed to defend—can themselves become attack surfaces, vulnerable to manipulation in ways that echo the threats they’re built to stop.

These developments raise uncomfortable but necessary questions. What happens when organizations without deep security expertise begin relying on AI to identify threats? Could they misinterpret results, act on false alarms, or overlook critical gaps? And how much transparency should we demand from tools that are shaping the very foundations of digital safety? Without rigorous oversight, we risk mistaking automation for understanding.

The truth is, AI can make defenders faster and more capable—but it can also make them complacent. If junior security professionals are trained only to audit what machines produce, we risk losing the very expertise we need to respond when those machines fail. Even now, many AI-generated vulnerability reports contain hallucinations—errors that mimic accuracy, but drain resources and undermine trust.

We shouldn’t shy away from AI in cybersecurity. But we also shouldn’t mistake automation for assurance. These systems can support us, but they can’t replace thoughtful governance, healthy skepticism, and the human judgment that resilience ultimately depends on.

Actionable Insights:

Machine Learning Software Engineer New Grad | Full time | NY, NY | Hybrid

Internship Opportunity | Internship | Year-round, Fall | In person

Low code/no-code app builder | Internship | Year-round | Hybrid

Entrepreneurship Center Brand Ambassador | Part Time | Year-round | Hybrid

Have a job or opportunity to share? Post for free here and we will highlight a few each week!

10 quintillion

bugs fill the planet—outnumbering us 1.4 billion to one, a reminder that the smallest lives often shape the biggest systems.