Artificial intelligence (AI) offers a clear advantage in fortifying cybersecurity for organizations and internet users globally. This becomes increasingly vital as malicious actors persist in targeting and exploiting crucial software and systems. The continuous development of AI has the potential to transform cybersecurity landscapes, with responsible AI deployment paving the way towards a more fortified digital infrastructure.
The journey towards achieving this goal necessitates robust partnerships between the public and private sectors and avenues for innovation, such as DARPA’s AI Cyber Challenge (AIxCC) Semifinal Competition, set to take place at DEF CON 32 in Las Vegas from August 8-11. This competition stretches over a two-year period, rallying AI and cybersecurity experts to create novel AI tools aimed at securing prominent open-source projects.
Initiated as part of DARPA’s commitments at last year’s Black Hat event, the Semifinal Competition follows the White House’s anniversary of voluntary AI pledges. This collaborative effort between industry and government aims to advance responsible AI practices. Google stands firm in its support of AIxCC competitors:
- Google Cloud resources: Google is empowering AIxCC teams by granting $5,000 in Google Cloud credits per qualifying team, with a total cap of $1 million. These credits can be utilized for eligible Google Cloud services, including Gemini, allowing participants access to Google Cloud AI and machine learning models to innovate and address various challenges during the competition. Encouragement is extended to leverage incentive programs like the Google for Startups Program, covering Google Cloud expenses up to $350,000 for AI startups.
- Cybersecurity experts: With a rich history of leveraging AI for security measures, Google is well-positioned to share insights with AIxCC participants. From advanced threat detection to safeguarding Gmail users against phishing attacks, Google’s security experts will offer guidance on projects, infrastructure, and establishing evaluation parameters. A guide highlighting AI’s potential to enhance Google’s open-source OSS-Fuzz technology has been released for AIxCC participants.
- AIxCC experience at DEF CON: Google will be present at the AIxCC Semifinal Experience at DEF CON, featuring AI-focused hacker demos and interactive showcases of Google’s AI technology. Attendees can explore Google Security Operations and SecLM on Chromebooks at the Google booth, engaging with Google security experts in technical dialogues.
AI Village at DEF CON 32
Google’s commitment extends to supporting the AI Village at DEF CON, offering insights into security and privacy concerns in the realm of AI. A $10,000 donation from Google will aid the AI Village’s efforts, with Pixelbooks provided for workshop attendees to delve into the fusion of AI and cybersecurity. Following last year’s triumph, wherein Google participated in the AI Village red teaming event, security professionals engaged in tackling AI security challenges.
The forthcoming AIxCC Semifinal Competition is anticipated to yield groundbreaking innovations that can be instrumental in fortifying the software infrastructure we rely on. Meanwhile, delve deeper into Google’s AI security initiatives through the Secure AI Framework and the recently unveiled Coalition for Secure AI.
https://blog.google/technology/safety-security/google-darpa-ai-cyber-challenge-support/
