Google froze its open source bug bounty on October 1, 2026. It stopped taking new product vulnerability reports because automated, mostly invalid AI submissions buried its reviewers. Reports filed before that date are still being handled. Google says it will share an update in the first quarter of 2027.
That is the whole Google open source bug bounty story in four sentences. The rest of this piece explains what is actually closed, what still works, and why this is bigger than one Google program.
What is the Google open source bug bounty?
It is the Open Source Software Vulnerability Reward Program, usually shortened to OSS VRP. Google launched it in 2022. It pays researchers who privately report security flaws in software Google releases as open source, such as Go, Angular and Protocol Buffers, according to Help Net Security.
The program also covers repository settings and supply chain components. Rewards depend on how important the project is. Google sorts projects into four tiers, from OT0 (flagship) to OT3 (low priority).
Think of the Google open source bug bounty as a tip line for the code that half the internet quietly runs on.
What exactly did Google freeze?
Google stopped accepting new product vulnerability reports through the Google open source bug bounty (OSS VRP) as of October 1, 2026. The official rules page says so directly, and you can read Google’s OSS VRP rules page yourself.
Here is what that means in practice:
- New product vulnerability reports: closed.
- Reports filed before October 1: not affected, still in the queue.
- Cloud VRP: may still accept product bugs for some Google Cloud repositories that affect Cloud products.
- Patch Rewards Program: still open. It pays for security improvements to Google’s open source projects, not for bug reports.
- Reward table: lists no payout amounts for product vulnerabilities in any of the four tiers.
Google says it will keep reworking its open source bug bounty and promises an update in Q1 2027. Notice the wording. The company says “pause,” not “shutdown.” Still, a pause with no reward amounts listed feels a lot like a shutdown to anyone who was hoping to get paid next week.
One caution. Google’s wording targets product vulnerabilities. We would not assume every other report type is open. Check the rules page before you spend hours writing anything up.
Why did Google freeze its open source bug bounty?
Google blames a flood of automated submissions that are almost all invalid. In an official X post, the company said the pause comes from a rise in automated submissions, the vast majority of which are not valid.
TechCrunch reported the same pause and noted that Google’s own program site announced it. Reporting from Tom’s Hardware, relayed by other outlets, adds that engineers and maintainers were swamped by thousands of weak reports. Many claimed to find bugs that turned out to be hallucinations or flaws nobody could exploit.
Here is the part that makes this hard to fix. A bug report costs a bot a few seconds and almost no money. Checking that report costs a human engineer real time. One person can fire off a hundred reports in an afternoon. A reviewer cannot clear a hundred in the same afternoon.
Every fake report still needs a judgment call. Meanwhile, the real bugs wait in the same queue.
AI is not the villain here. It is a strong tool, and we have seen it get good at security work. OpenAI’s Astra model reaching the “Critical” cybersecurity threshold [[IL2]] is one example of how capable these systems have become. The problem is people pasting unverified model output into a form and hoping for a payout.
Is Google the only company pulling back?
No. The Google open source bug bounty freeze is the latest, not the first. Several programs have tightened up or closed in 2026, and the reasons overlap.
| Program | What happened | Reason stated |
|---|---|---|
| Google OSS VRP | Product vulnerability reports closed Oct 1, 2026. Update due Q1 2027. | Rise in automated submissions, most invalid |
| Intel bug bounty | Paid program (up to $100,000 per flaw) replaced by a no-reward disclosure program on Intigriti, Sept 2026 | No official reason given |
| curl | Ended its bug bounty earlier in 2026 | Maintainers cited a flood of AI-generated reports |
| Internet Bug Bounty (HackerOne) | Paused new submissions | Rising AI-assisted discovery volume |
| Linux kernel | Dropped support for some older network drivers | Reported as a response to false AI-generated bug reports |
A fair warning on the Intel row. Tom’s Hardware reported that Intel gave no reason. Linking it to AI slop is a reasonable guess, not a confirmed fact. We list it because the timing fits, not because Intel said so.
Google’s case is different. Google named the cause itself.
Is this really an AI problem, or a bounty design problem?
Both. Our view: the pay-per-report model behind the Google open source bug bounty was always easy to game, and AI made gaming it nearly free.
Before language models, a junk report still cost the sender some effort. Now it costs nothing. When the cost of spam drops to zero, spam wins. The honest fix is verification. Programs will likely ask for a working proof of concept, a reproducible test, or a track record before a report ever reaches a human.
The attacker side is moving fast too. We covered how AI agents breached 440 PaperCut servers [[IL1]] across 395 organizations in 48 countries. Defenders are being outpaced on speed, and now they are also drowning in noise. That combination should worry anyone who relies on open source.
What can security researchers do right now?
Use other channels and bring proof. If you planned a report for the Google open source bug bounty, here is a sensible order of operations.
- Read the rules page first. It is the only source that tells you what Google accepts today.
- Check Cloud VRP. Some Google Cloud repositories that touch Cloud products may still take product bugs there.
- Look at the Patch Rewards Program. Writing a real security fix can pay when a report no longer does.
- Try Google’s other bounty programs. Google points researchers toward its other VRPs, including Cloud and AI programs.
- Reproduce everything yourself. If an AI tool found the bug, run the exploit path by hand. If you cannot reproduce it, do not submit it.
- For non-Google projects, follow the project’s own security policy. Many maintainers prefer a private email over a form.
This matters for honest researchers most. The people who get hurt by AI slop are the careful ones. Their good reports now sit behind a closed door.
What does the freeze mean for developers in Pakistan and beyond?
It raises the bar for newcomers. From our desk in Pakistan, we know many young developers treat bug bounties as a first side income. That path just got narrower now that the Google open source bug bounty is frozen.
We think the smarter move is to build a reputation on quality. One verified, well-written report beats fifty copy-pasted ones. Patch Rewards is also worth a look, since it rewards fixing code rather than flagging it.
For readers in the US, UK and Australia who maintain open source projects, the lesson is practical. Add a clear security policy. Ask for reproduction steps. Decide now what you will do when the slop reaches your inbox, because it will.
Will AI slop end bug bounties for good?
Probably not, but the open-door version is in trouble. The Google open source bug bounty is the most visible casualty so far. We expect more programs to add stricter proof rules, invite-only tiers, or reputation gates. AI was supposed to make vulnerability research easier. It did, for everyone, including people with no skill and no scruples.
Our own experience with AI-assisted security writing at Innvobyte backs this up. The tools are helpful when a human checks the output. They are dangerous when nobody does. Anthropic’s own threat intelligence report on Claude misuse [[IL3]] shows the same pattern from the platform side: capability is rising, and so is abuse.
What to do next: bookmark the Google open source bug bounty rules page and check it again in Q1 2027. If you run a project, write your security policy this week and require a working proof of concept. If you hunt bugs, spend the pause learning to write one airtight report instead of ten weak ones.
FAQ SECTION:
Q1: Why did Google freeze its open source bug bounty? A1: Google paused product vulnerability submissions to the OSS VRP because of a significant rise in automated submissions, most of them invalid. Reviewers and open source maintainers were swamped with AI-generated reports that described bugs that were hallucinated or could not be exploited. Google says it will give an update in the first quarter of 2027.
Q2: Is Google’s open source bug bounty permanently closed? A2: No. Google calls it a pause, not a shutdown. As of October 1, 2026, it no longer accepts new product vulnerability reports through the OSS VRP. The company says it is reformatting the program and has committed to an update in Q1 2027. Nothing has been announced about what the new version will look like.
Q3: Are all Google bug bounty programs paused? A3: No. The freeze applies to product vulnerability reports in the Google open source bug bounty (OSS VRP) only. Google still points researchers to its other programs, including Cloud and AI VRPs, and to the Patch Rewards Program. Some Google Cloud repositories may also still accept product bugs through Cloud VRP when they affect Cloud products.
Q4: What happens to reports submitted before October 1, 2026? A4: Reports filed before October 1 are not affected by the pause. Google is still processing them under the existing rules. The freeze only blocks new product vulnerability submissions made on or after that date. If you have an open report, keep an eye on its status in your submission portal.
Q5: What is AI slop in bug bounty reports? A5: AI slop means low-quality, machine-generated vulnerability reports sent without verification. They often look professional, with technical language and file references, but describe flaws that do not exist or cannot be exploited. Each one still takes a human reviewer time to check, which is why they overwhelm security teams.
Q6: Can I still get paid for finding bugs in Google open source projects? A6: Not through OSS VRP product vulnerability reports right now. The reward table lists no payout amounts for them in any tier. You can still pursue the Patch Rewards Program, which pays for security improvements, and Cloud VRP for qualifying Cloud repositories. Always confirm current rules on Google’s rules page first.
