How we found 24 Android vulnerabilities using our open source AI security agent
GitHub Security Lab’s AI taskflows uncovered 24 Android bugs, including OsmAnd location tracking.
GitHub Security Lab described open-source Taskflow Agent workflows used to audit Android applications and said researchers have reported 24 vulnerabilities. The prompts separate mobile entry points and direct the model toward classes such as confused deputy and insecure broadcasts. In OsmAnd, which has more than 10 million downloads, an exported MapActivity accepted intent extras meant for an internal service, allowing another app to silently import settings and potentially track location. The post does not say the flaws are being exploited; running the taskflows requires GitHub Copilot.
- Open-source GitHub Security Lab taskflows found 24 Android vulnerabilities.
- Workflows separate mobile entry points and check known mobile bug classes.
- OsmAnd’s exported MapActivity accepted intent extras for silent settings import.
- A malicious app could use that path to track the device location.
- Running the audits requires GitHub Copilot and can consume many tokens.
- Vendors
- GitHub
- Organizations
- GitHubGitHub Security Lab
From the original article
10 min read at github.blog
A look at the targeted AI taskflows behind these findings, the critical Android bugs they uncovered, and how to run the same open-source agent on your own app. The post How we found 24 Android vulnerabilities using our open source AI security agent appeared first on The GitHub Blog.
By Kevin Stubbings · github.blog
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.