Mountain View, CA – July 16, 2025 — As the cybersecurity world prepares for its biggest conferences of the year — Black Hat USA and DEF CON 33 — Google is taking center stage with a sweeping announcement: a new era of cybersecurity is here, powered by cutting-edge AI and driven by collaboration. Today, Google shared breakthroughs in AI-driven security tools, unveiled industry-shaping partnerships, and emphasized its commitment to securing the digital world through innovation, responsibility, and cooperation.
With the evolving digital threat landscape, cybersecurity professionals face increasingly sophisticated attacks. Google believes the key to turning the tide lies in leveraging AI to give defenders the upper hand. Through new agentic capabilities, advanced platforms, and public-private collaboration, the tech giant is setting a bold tone for “A Summer of Security.”
Agentic AI: The Future of Threat Detection is Here
One of the most exciting developments comes from Google’s AI agent known as Big Sleep, developed by Google DeepMind and Project Zero. Originally designed to detect unknown software vulnerabilities, Big Sleep made headlines last year when it discovered a real-world flaw before it was exploited.
Now, it’s breaking even more ground. Powered by threat intelligence, Big Sleep recently uncovered CVE-2025-6965, a critical SQLite vulnerability that had been secretly targeted by malicious actors. This marks a cybersecurity milestone — the first known instance of an AI agent preemptively thwarting an exploit attempt in the wild.
Beyond securing Google’s ecosystem, Big Sleep is now being used to bolster the safety of open-source software across the internet, helping defenders scale their efforts and address vulnerabilities faster than ever before. Google also released a white paper outlining its responsible approach to AI agent development, ensuring safety, human oversight, and transparency remain at the forefront.
“These agents are a game changer,” Google said in a statement. “By allowing defenders to focus on the most complex threats, we’re amplifying their impact and helping to redefine modern cybersecurity.”
Next-Gen Security Tools Take the Stage at Black Hat and DEF CON
Google is also introducing a suite of AI-powered tools designed to revolutionize incident response and threat detection:
- Timesketch with Sec-Gemini Integration: Google’s open-source forensic analysis platform now leverages agentic AI to perform initial forensic investigations autonomously. This drastically reduces time-to-response and allows analysts to focus on complex tasks. A live demo will be available at Black Hat USA, booth #2240.
- FACADE (Fast and Accurate Contextual Anomaly Detection): An insider threat detection system that’s been operational within Google since 2018, FACADE processes billions of daily events to flag anomalous behavior, without relying on data from prior attacks. Attendees at Black Hat will get an exclusive look at how it works.
- DEF CON GENSEC Capture the Flag (CTF): Google and Airbus are teaming up for a hands-on challenge at DEF CON 33, where participants will partner with AI assistants to tackle cybersecurity puzzles, showcasing how human-AI collaboration can elevate defense capabilities.
Public and Private Collaboration: Building a Safer Internet Together
Recognizing that no one can tackle global cyber threats alone, Google is deepening its commitment to cross-sector collaboration:
- The company is donating valuable Secure AI Framework (SAIF) data to the Coalition for Secure AI (CoSAI) — an initiative aimed at responsibly advancing AI adoption in cybersecurity and software supply chain protection.
- Google also spotlighted the upcoming finale of the AI Cyber Challenge (AIxCC) with DARPA at DEF CON 33, where top innovators will unveil AI tools designed to identify and patch vulnerabilities in major open-source projects.
“We’ve always believed in AI’s potential to make the world safer,” the company noted. “Now, we’re seeing that potential being realized — and it’s only just the beginning.”
Looking Ahead: Securing the Digital Future with AI
Google’s summer of security marks a critical moment in cybersecurity history. Through strategic advancements in agentic AI, next-gen tools, and industry-wide cooperation, the company is laying the foundation for a future where digital safety is scalable, predictive, and resilient.
As threats grow in scale and complexity, so too must our defenses. With responsible innovation and collective action, Google envisions a digital world that is not only more prosperous but also more secure for everyone.
Media Contact:
Google Communications
press@google.com
Read more: How Agentic AI is Redefining Global Capability Centers: Top 10 Global Use Cases to Watch in 2025