Google Pauses Open-Source Bug Bounty After Surge of Invalid Automated Reports
Google has paused its Open Source Software Vulnerability Rewards Program. The pause took effect on October 1, and Google said it will provide an update in the first quarter of 2027. TechCrunch
The program rewarded researchers for finding vulnerabilities in Google's open source software. For researchers with in-flight work or ongoing interest, Google encouraged participants in the paused program to consider its other bug bounty programs.
Google said the pause was due to a significant rise in automated submissions, the vast majority of which are not valid. That language points to volume as the operational problem, not a single incident or a change in scope for open source security itself.
The freeze is narrow in one important respect. It applies specifically to product vulnerability reports. Supply-chain disclosures submitted under the OSS VRP remain open despite the freeze. Times of India Google attributed the freeze to a rise in invalid AI-generated submissions. Tom's Hardware
The distinction matters for triage. Product reports typically require reproduction, severity assessment, and coordination with maintainers. Supply-chain disclosures follow a different workflow, focused on dependency metadata and downstream propagation. Keeping the second channel open preserves a path for the class of reports where automation and scalable analysis add the most value.
This is not Google's first adjustment to how it pays for external security work. Google published reward criteria for reporting bugs in AI products to help its bug hunting community test AI safety and security. Google Security Blog Google later reported that it received over 150 bug reports in the first year of its AI bug bounties, paying over $55,000 in rewards. Google Security Blog More recently, Google launched a patch rewards program for OSV-SCALIBR, its open source tool for finding vulnerabilities in software dependencies. Google Security Blog
Taken together, that history describes a vulnerability rewards operation that expands into new areas while pruning channels that stop working. New targets get criteria and incentives. Tooling gets its own patch incentives. A noisy queue gets frozen.
The broader context here is familiar to anyone who has operated a bug bounty queue. The cost of generating a plausible-looking report has fallen sharply. The cost of validating it has not. A reviewer still needs to reproduce the build, check the vulnerable path, assess exploitability, and decide on severity. When the submission rate rises and the validity rate falls, the queue consumes engineering time without improving security posture.
In my view, the pause reads less as a retreat from open source security than as an admission that the current incentive design does not survive automated submission at scale. Pay per valid report works when finding candidates is hard. It breaks when candidates are cheap to manufacture and expensive to reject. The fix is unlikely to be simply paying more or hiring more triagers.
Looking at what this means for researchers, the near-term signal is straightforward. Low-effort, high-volume submission strategies now carry a clear penalty. They do not just fail to pay. They risk closing the programs they target. Researchers who invest in reproduction, proof of exploitability, and clean scoping will remain valuable. The challenge for platform operators is to make that value legible in the submission process itself, through stronger proof requirements, reputation weighting, or tool-assisted pre-validation.
Worth flagging is the split Google chose. By freezing product vulnerability reports while leaving supply-chain disclosures open, Google preserved the more structured, machine-checkable input and paused the more judgment-intensive input. That is a pragmatic way to protect reviewer bandwidth. Over the longer arc, that split may point to where bounty programs go next, toward tighter schemas, reproducible artifacts, and automated pre-checks before a human ever looks.
