Google's AI global copyright infringements at scale needs to stop

[Whitey]

Thanks @goodoldweb. This is one of the most important and grounded posts I've seen in recent months, so i extracted it from the thread and started an OP:

Webmasters, it’s time to face the truth: no one is going to protect our content from Google’s AI Overviews—except us. Every day, original work is being lifted, repackaged, and served back to users with no or very little attribution, and no traffic returned. This isn’t curation. It’s exploitation.

Every one of us, in our own countries, has copyright laws that protect our work. We don’t need to wait for a global solution—each of us can push back, starting at home.

So let’s stop waiting. Each of us has the power to act within our own legal systems. File complaints. Demand accountability. Push your local authorities and copyright bodies to take this seriously. Almost every country has a copyright authority with a website and a complaint form—start there. The more of us submit complaints, the harder it is for them to ignore. Google isn’t above the law—and it’s time we remind them of that.

I understand some webmasters fear filing takedown requests might risk their sites being de-indexed by Google, but staying silent only lets the problem grow. When filing complaints with authorities, it’s important to mention this fear—so they understand the full impact on webmasters and can take it seriously.

And just as important—we must all demand real tools to block AI from scraping our content. This shouldn’t be opt-out. It should be opt-in. Our work should not be fair game just because it’s online. Consent must be the default.

Do not delay. Do not wait for someone else to lead. Every day we stay silent, more of our work is taken without credit, without permission. If we don’t act now, we’re handing over our rights without a fight.

In Australia, our copyright law includes ‘fair dealing’ exceptions—but they’re limited to specific uses like research, criticism, or news reporting. Simply scraping or repurposing our content without permission doesn’t qualify. That means we have strong legal protections—if we choose to use them.

Today, we filed an official complaint with the Australian Competition and Consumer Commission (ACCC), and also with the Australian Copyright Council, to hold Google accountable—and I encourage you to take action in your own countries too.

P.S. I wish this website would take a more prominent role. It is a gathering place for webmasters from all over the world and, in a way, represents webmasters everywhere.

What’s playing out isn’t just a technical shift, it’s a structural one. The long-standing "value exchange" between content creators and platforms is eroding. AI Overviews increasingly reframe your original work into zero-click summaries, often without attribution, credit, or meaningful visibility in return. It's hard to see that as anything other than extractive.

In the Australian context, as mentioned, 'fair dealing' has well-defined limits, and broad repurposing for commercial advantage doesn’t comfortably sit within those. The underlying legal frameworks do exist in most jurisdictions, though they’re not always visible or enforced in the context of AI scraping and synthesis.

What’s also becoming clear is that the web’s feedback loop (user visits, engagement signals, backlinks) are being bypassed. As a result, traditional SEO strategies are starting to decouple from visibility outcomes. It raises a difficult but important question for many site owners: where is the line between exposure and exploitation?

Whether or not one pursues formal action, understanding the boundaries of copyright, fair use/fair dealing, and digital property rights seems more relevant now than ever. It’s no longer just about rankings, it’s about how the web sustains original creators.

A tactic that makes actionable notice more likely is to file complaints and cc your local politicians, authorities, ombudsmen, key journalists in your jurisdiction. Almost like an "Open Letter" that hit's the mark. The aggregated effort cannot be ignored by making those accountable to each other in the cc'd correspondence.

I'll be following up on this over the next few days.

[not2easy]

The founder and owner of this site set the Terms of Service. Contact Brett Tabke if you want to discuss changes, but you might want to see his May 20, 2025 post here: [webmasterworld.com...] before you ask.

[RubicCubed]

@saladtosser

While I can understand a forum's desire to remain impartial, I also agree with you that such impartiality also has the potential to shield a company that has recently been handed it's second antitrust crime conviction in a USA Federal court. At the very core of being a webmaster is publishing, and I agree with you that the very essence of our existence is being threatened by Google. I would also agree this and other forums should review their current policies to better reflect the communities they represent and if their existing policies conflict with the overwhelming needs of the communities they represent.

So why don't we band together and start a class action?

Without violating #26 claims of action, have you considered reaching out to Chegg's attorneys to see what they say about your idea? Susman Godfrey L.L.P. is no fly by night law firm, having achieved a $787 million settlement for their client against Fox News, and they do handle class action lawsuits which the Chegg lawsuit could serve as the foundation for. If you view the full Chegg -vs- Google lawsuit [storage.courtlistener.com...] (PDF) at the bottom is the contact information for the attorneys representing Chegg.

Edit: I had not seen not2easy's post before I submitted mine. That terrible to hear the bot activity is so extreme it could eventually force the closure of this site. More joys of the AI age with everything being scraped by everyone. :(

[Whitey]

Edit: I had not seen not2easy's post before I submitted mine. That terrible to hear the bot activity is so extreme it could eventually force the closure of this site. More joys of the AI age with everything being scraped by everyone.

FWIW - I reached out to Brett with a post about a possible tech solution to these bot attacks. Our sites were hammered last week with multi geo sophisticated attacks and we struggled to stay online, but our CTO organized a potential remedy after programmatic IP blocking became unworkable.

[webmasterworld.com...]

Please help him if you can.

[Whitey]

Thanks @RubicCubed, @not2easy, both posts reflect the strain many of us are feeling, not just from legal blind spots, but from technical ones as well.

The conversation around forum policy, legal pushback, and technical defense needs to stay active. I’m encouraged to see names like Susman Godfrey being mentioned, this adds real substance to the legal angle. Meanwhile, the scraping issue isn’t hypothetical anymore. It’s driving existential decisions for smaller publishers and forums alike.

We’ve entered an era where site owners, creators, and aggregators are becoming invisible labor for LLMs, stripped of credit, traffic, and often purpose. If WebmasterWorld stands for publishers, then evolving the TOS to reflect today’s threats isn’t activism, it’s survival strategy.

My 2c: continue surfacing viable solutions—legal, technical, strategic, so we stay within the rules here but push the conversation forward constructively. Brett’s post [May 20, 2025 mentioned above] acknowledged the urgency. We owe it to this community to work with him in shaping a path forward.

[Micha]

Edit: Oops, I didn't see the second page of the thread, so I didn't say anything ;)

[edited by: Micha at 5:03 pm (utc) on May 25, 2025]

[not2easy]

You can always chat via sticky mail to share your contact information with others so you can then decide what format (chat, email, website, platform) you want to use. Heck, a noindexed, free WP.com site* can be used for the purpose but it is not allowed here due to legal concerns.
_{And no, Google does not support WebmasterWorld forums, that's why it says any company.}

*(AFAIK- but WP.com may also limit organizing against a company)

[saladtosser]

Whitey have you considered putting honeypots on every form and 301 the gits to another site? I put about 5 separate traps on everything that can be submitted on my site from login to search features and more and they have helped loads, depends if you have that sort of functionality on your site I guess!?

[Whitey]

FYI

@brett_tabke - In case it helps an internal message from my CTO regarding attacks that brought us down this last week;
Based on recent bot attacks from China that have impacted our servers, we need to implement some general protection, rather than keep blocking ip addresses. The plan is to install Crowdsec & Zenarmor, which compliment each other, to combat these attacks.

Both are paid services, but both offer free options. I suggest we implement both, but only pay for Crowdsec initially. If necessary, we can also upgrade to a paid version of Zenarmor, which ism’t much more than Crowsec.
[webmasterworld.com...]

[tangor]

The problem with "organizing" is the MANY different geo-locations involved. Each has laws of their own that might be similar in intent but not the same IN WORD or ENFORCEMENT, creating all kinds of conflicts. There is no single global entity that can address this topic for ALL webmasters, thus any action will occur at the nation/state level.

Additionally, there are nation/states who will never comply due to ideological and political differences so even if some fall in line dealing with AI overreach on copyright theft those other actors will turn a blind eye and continue such activity. There is no way to PREVENT that theft short of excising entire swaths of the internet.

Not likely to happen!

Looking at the templates suggested above: Won't work. TOO LONG. TOO WORDY. USERS DON'T CARE (they want the instant "free" gratification promised by the WWW...). USERS are not your target, heck, you can't live WITHOUT them! Targeting G merely gets you disappeared. Targeting AI Bots is whack-a-mole on steroids. AI is essential to nation/state investment, progress, and security, thus it will NOT go away even as it wrecks Creators with lack of attribution or pass-through. THEN AGAIN, how many websites are ACTUALLY Creators of New and Copyrightable Property? LOOK AND FEEL (a US copyright term) only goes so far, particularly if the actual content is rehashed from known and general knowledge that can't be copyrighted in the first place.

AI is a game changer in the absolute broadness of ignoring copyright and fair use (US term, though other countries have their version of same). THAT is the major obstacle, as well as the others listed above. Is there a solution? A decade or so of litigation MIGHT provide an answer. AT PRESENT the most obvious action for webmasters is to ban, deny, reject all AI intrusion at the likely cost of lost traffic and revenue.

[Whitey]

It’s clear we’re all grappling with a shifting ecosystem; legal ambiguity, platform dependency, technical strain and still trying to keep our communities and content afloat.

@saladtosser: Great idea on honeypots; our dev team has started implementing CrowdSec + Zenarmor, (per above) and early signals look promising.

@tangor: Your realism is appreciated. I see that you're right that the fragmentation of global enforcement complicates coordinated action, but when a platform like Google effectively is the internet in the Western World, providing 85%+ market visibility in search, that scale brings a unique level of responsibility and scrutiny. Targeting the right levers, nationally and platform-specifically, may still move the needle, even if slowly.

I agree, users crave instant answers, but we’re also seeing early signs of backlash as the consequences of generative overreach hit creators, site owners, and public trust. So the legal, technical, and strategic groundwork we're discussing now may turn out to be more timely than we think.

Let’s keep surfacing practical solutions here, whether it’s defensive tech like @saladtosser's, viable schema strategies, or sharing public-facing narrative templates that frame the issue without falling foul of TOS.

We’re stronger for having these threads

[RubicCubed]

The problem with "organizing" is the MANY different geo-locations involved.

There is truth to this statement. But from what I've seen the real problem is most want others to do the heavy lifting for them. Regardless of where we live, we all have elected representatives and regulators. In the spirit of #26, what has each one of us done individually to air our grievances with those who represent us? Just because there is no organized effort, or class action I can join, doesn't mean I am or will remain silent. This is my career, and I value it enough to defend it by speaking up to the powers that be.

Great idea on honeypots

Tarpits seem to be an anti-AI scraper solution some are using. It definitely won't help save server resources, but gives those who want to fight back against AI content theft something to consider.

[pcworld.com...]

[Whitey]

TL:DNR - Google’s AI Mode Patent Confirms the Threat: Copyright + Monopoly = Clickless Future

Google’s AI Mode patent confirms [patents.google.com...] what many feared: AI-driven search will summarize our content without attribution, clicks, or consent. As Michael King explains, [linkedin.com...] LLMs will generate synthetic queries, extract answers, and bypass the traditional link model, all while monetizing responses.

AI Mode is not just a UX shift, it’s a business model pivot at our expense. We feed the machine, and it shuts the door behind us. Monopoly and copyright theft imo

[Kendo]

Pushing points of law about copyright and extortion won't do much while the door is left open.

I haven't forgotten about this and will soon be able to provide an example of my bot solution as POC.

[Whitey]

TL;DNR: Press Gazette just published a detailed roundup on how major news publishers are either suing or cutting deals with OpenAI and Google over the use of their content in AI systems:
[pressgazette.co.uk...]

The article outlines a fragmented global landscape; some publishers (like Axel Springer and AP) are signing licensing agreements, while others (NYT, Le Monde, and others) are going to court, arguing unauthorized use of their copyrighted material to train LLMs and power tools like AI Overviews.

This underscores how Google is trying to sidestep a total legal blockade by selectively signing publishers to quiet dissent while the real problem; unauthorized scraping and content repackaging, remains unresolved for the rest of us. There’s no opt-in, no attribution, and certainly no compensation unless you’re part of the elite media club. Small and mid-size publishers, bloggers, and e-commerce operators are left to fend for themselves in this “AI land grab.”

If deals are happening, where’s the transparency? Where’s the shared standard or open call to participate?

This patchwork response should strengthen the alarm to everyone reliant on Google traffic. It’s a divide-and-conquer approach, and it’s working.

We need to be asking: Who’s negotiating for the rest of the open web and us? Or are we just data feed?

[RubicCubed]

Who’s negotiating for the rest of the open web and us? Or are we just data feed?

I feel these deals prove that intellectual property has value, whether it be the New York Times' published pages or ours. BTW, Amazon just entered into an agreement with NYT [cnn.com...] to train their AI too. Amazon, Google, OpenAI or whoever shouldn't be allowed to pick and choose who they want to pay to use their content in AI. Licensing intellectual property should be mandatory, and I think the Chegg lawsuit will help in those regards UNLESS it gets settled before trial.

[goodoldweb]

I believe the most effective way to handle AIO content theft right now is to allow bots (any bots) access only to article excerpts, with full access granted only after passing a human verification test. In fact, I've already begun implementing this solution across several of our websites.

Sorry Google you can't have my work and expertise stolen and published without clear attributions.

Hopfully alot more publishers will join the fight by implementing similar solutions.

[Skips]

I feel the frustration many of us feel. My own traffic is down ~20% week-on-week, while bot traffic is 50 x higher. That says a lot. Google’s AI Overviews seem to take our content, repackage it, and serve it back to users with no or little attribution and traffic in return. That breaks the long-standing deal: we give content, search engines send us visitors.

That said, I think a legal fight will be tough. Once content is “repackaged” by AI, it becomes a blurred summary of many sources. Unless it’s clearly lifted from a unique piece, proving copyright infringement is difficult. The bigger issue here is ethics, and unfortunately, Google doesn’t seem too concerned with that.

Ironically, I raised this with ChatGPT about a year ago, it acknowledged the importance of attribution and not long after, I started seeing leads from it. Google? Not the same story.

The real danger for Google is that if webmasters get nothing in return, many will start blocking or limiting access to their content. That would hurt Google’s AI over time, because its knowledge would become outdated fast. But it looks like they need to feel that pain before they act.

That’s why I believe we should focus on technical solutions, not just legal ones. For example:

- Allow only partial or outdated content to be crawled
- Reserve fresh or valuable content for human visitors
- Use robots.txt and user-agent filters (like blocking Google-Extended)
- Monitor and limit excessive bot traffic

We definitely need to proactively defend our work, but technical action might be our fastest, most effective path. What’s everyone else seeing or doing in response technically-wise?

[Kendo]

I am finding search annoying. Now that both Google and Bing are using AI results, it is slowing down results and I am finding the information unreliable and at times inaccurate.

For a technical action to have an effect everyone needs to use it. But I doubt if that will happen. If anything many might opt to sit back and watch their ranking increase while we reduce our search engine fodder.

But I am still working on my solution. A preview won't be long now.

[mhansen]

Sorry Google you can't have my work and expertise stolen and published without clear attributions.

I don't think attribution is enough any longer. My own personal belief is that we need to forget about search and free web as we know it, and make make all bots pay a fee each time they come visit to scrape and download. Similar to Googles own API usage for maps or the way hosting requires payment for data usage, etc. Someone with widespread usage like CloudFlare (Or Google itself) could easily implement a solution like this and publishers should be paid upfront, in order to have their copyrighted materials scraped.

One of my sites - yes, it's still semi-popular in Google, but the spread between clicks and impressions is quickly widening, even with many top 3 organic placements, has seen 201,000 bot visits this month and billions of bytes of data harvested to feed their machines. Since this content is to be used over and over in their AI/LLM content and they earn who knows how much from the ads they inject, I think a fair use payment is a good tradeoff. Pay us on the front end, like you do with large publishers, or you cannot have the copyright-protected content.

[Kendo]

paid services to prevent DDos attacks

If using Windows Server you don't need any of that.

Microsoft provides a module for Dynamic IP Restrictions.

The Dynamic IP Restrictions (DIPR) module for IIS 7.0 and above provides protection against denial of service and brute force attacks on web servers and web sites. To provide this protection, the module temporarily blocks IP addresses of HTTP clients that make an unusually high number of concurrent requests or that make a large number of requests over small period of time.

See [learn.microsoft.com...]

To Install

From the Select Role Services screen, navigate to Web Server (IIS) > Web Server > Security. Check the IP and Domain Restrictions check box and click Next to continue.

[Whitey]

Excellent perspectives here. The writing’s on the wall: unless you’re a mega-publisher with lawyers and lobbyists on speed dial, you’re not getting a seat at the AI licensing table. We’re the training set, not the negotiating party.

What strikes me is how fast this “AI land grab” has normalized. Google signs a few high-profile deals, launches AI Overviews globally, and everyone else, independent publishers, medium and small sites, bloggers, regional media, just watches their traffic vanish into the black box. And let’s be real, most users don’t click through anymore when the summary is handed to them upfront.

I agree with @Skips and @mhansen, we need collective technical responses now, not just legal ideals. A few things our crew is experimenting with:

•Serving partial or obfuscated content to bots

•Blocking Google-Extended and other aggressive crawlers

•Creating bot traps to monitor & fingerprint unauthorized access

•Working on story “wrappers” for AI: blocks of value-added content that don’t lift cleanly into LLMs

And I think it’s time to seriously consider federated action: a decentralized alliance of small/indie publishers that sets terms, builds tech together, and forces transparency. But the numbers participating need to be far greater than the engagement we have here.

Because if we don’t define the rules, we’re just feedstock for someone else’s platform.

I wonder what technical or collaborative efforts are others seeing out there.

[Skips]

For years, Google told webmasters: use structured data! Add schema markup! Optimize for rich snippets — it’ll help your visibility!

...and we did. We made our content cleaner, more machine-readable, and more extractable.

Now? Google is using that same structure - the very one we were told would help us - to power AI Overviews and answer boxes that strip our traffic and give users everything they need to know without ever clicking through.

It seems this was never about visibility - it was about making our content easier to harvest.

Some could pursue the let's just rank top 3 strategy, but that misses the point. AI doesn’t just learn from top 3 pages. It learns from the whole web. And when smaller sites start realizing that Google takes their answers but sends them nothing back, they’ll start blocking Googlebot. Quietly at first, then en masse.

It won’t happen in one day, but it will snowball. By the time Google notices the quality of its results degrading, users will have found alternatives.

Google’s biggest risk now isn’t competitors, it’s the death of trust. The small web should wake up.

[BigKat]

Some could pursue the let's just rank top 3 strategy

We have #1 ranks that now get nearly no traffic. Being cited, which we are, doesn't generate much traffic either. What Google is doing is like a TV network buying all the other networks in the world and slowly cutting those programs so only the shows/movies they produce are shown. People would freak out if this happened, but that's exactly what Google is doing to the web.

[Kendo]

Monitoring trends and creating startups based on our activity has never new to them, nor is favouring paying advertisers and investors. So if you are in one of their industries you are doomed, because lets face it, which industry do they not want to get into. Or better still, what do they not want to do to make a profit?

[Skips]

I'm seriously considering serving generalized, keyword-relevant content to all bot traffic instead of real-time calculated results. Not to game rankings, but because generating full dynamic content for bots is creating massive server load with over 500k bot hits daily I've been getting over the past few days. I know this technically qualifies as cloaking, but I see it more as resource prioritization: the keywords and content remain relevant, just not computationally heavy. Until bots start bringing actual value (like traffic or conversions), I'm not convinced they should have the full content experience meant for human users.

[goodoldweb]

The way forward:

Complete a human verification to unlock the full content.

Everyone else just gets a glimpse.

The more of us implementing this solution the faster we can turn that ship.

[Kendo]

They probably already have all of our current content so anything that we do now will only protect new content. Perhaps the recent spike in spidering is their last ditch effort to get what they can before we close the door.

Redirecting them to "summary" pages might reduce their existing content.

[tangor]

Just a reminder on the bot front: Whitelisting.

"These and NO OTHERS allowed"

Won't stop it but will quickly reveal who doesn't play nice. Helps put "winners and losers" choice back into the marketplace.

Cloaking and Summary pages, however, breaks the current paradigm (as intended!) but will have more consequence on webmasters rather than AI bots/g etc.

The part that is not being said out loud is that USERS don't care---and won't care---even if aware. They are insatiable, greedy (free), and instantaneous. Without allies that count any "war" will be difficult to prosecute, much less win.

</doom&gloom="thatsjustthewayitis">

[Kendo]

My bot solution is very close to being ready for road testing.

Bots that toe the line by agreeing to pay for usage are entitled to nominate sites by category/s and review a list of available sites to crawl which shows how many pages are available and each site's tariiff, for example $0.01/hit. This way they can exclude any sites that they consider unrelated to their agenda or too expensive.

But while this list is not public, bots can, by creating an account, get a list of all member sites to "do evil"!

I am not sure what we can do about that. Any ideas?

[Whitey]

Totally agree with the direction this thread is taking; we’re not just seeing a traffic decline, we’re witnessing a fundamental shift in who captures value from content.

Google’s move is less about search and more about owning the outcome; AI answers, knowledge panels, scraping and summarizing, all while cutting out the publishers that built the web in the first place. Even top rankings now barely move the needle. Citations are worthless if they don’t drive visits.

@Kendo’s idea of a bot pay-per-crawl whitelist isn’t crazy; it’s overdue. We’ve licensed human content for decades; why should bots get a free ride, especially when they’re not delivering reciprocal value? If AI companies want to crawl at scale, they should pay for access or accept summary-level content. Simple, imo.

In the meantime, it makes sense to serve lean, cached, or static content to bots. That’s not cloaking, that’s resource triage. 500K+ bot hits a day isn’t sustainable, especially for sites running real-time outputs or expensive API calls. Human gating might feel aggressive, but if that’s what it takes to survive, so be it.

There’s no cavalry coming. The future is in owning relationships directly with email, apps, social, even paid or walled communities etc etc. Playing the Google game passively and waiting for roll back to the past is no longer a strategy. It’s surrender.