Philip Martin, Internet Erasure
By Philip Martin, Senior Caseworker at Internet Erasure Ltd [1]
The Right to be Forgotten is one of the most significant privacy rights created in modern times.
Like most legal services, Right to be Forgotten applications and processes are being changed by Artificial Intelligence.
AI can guide and improve individual submissions, but in this article, we will highlight why professional representation still dramatically improves the likelihood of success.
Established following the landmark Google Spain decision and now reflected in Article 17 UK GDPR and EU GDPR, the Right to be forgotten allows private individuals to request search engines to stop displaying links to information about them when their name is searched, if the content is outdated, irrelevant, inaccurate, inadequate or disproportionately damaging.
The Right to be Forgotten requires a careful balancing exercise between an individual’s data protection and privacy rights (including rehabilitation in criminal cases), and the wider public interest in continued access to that information. Individual applications are considered on their own merits. Important considerations before a decision can be reached to remove content from search results include severity, recency, current relevance, ongoing public interest or public protection, and the impact of continued indexing on the individual.
Restoring privacy within the framework of law is not a form of censorship. In most cases it does not require publishers to remove content, nor does it rewrite the public record. Instead, it limits the appearance of links within search results generated by searches for an individual’s name. The publications and related content usually remain available through other search terms and by direct access.
Negative search results appearing when an individual’s name is searched can be a significant barrier to employment and other opportunities, can prevent the renting or purchase of a home and can significantly affect personal relationships.
The ripples of adverse media often affect school-age children too, when they are taught to search for their parents online as part of classroom IT exercises. In an age of almost permanent digital records, it is difficult to understand why this remains common practice.
The oldest publication we have removed from search engines dates from 1996. This single negative article had affected the individual for almost 30 years. They described life with the article continually appearing as “driving with the handbrake on”.
The widespread availability of AI and its recent capability improvements have now prompted a new question…
If AI can now research legislation, analyse case law and draft persuasive legal submissions within minutes, has specialist representation become less valuable?
To answer this, we need to understand what a modern Right to be Forgotten application involves.
There are two distinct disciplines involved in achieving clear search results: legal work and technical work. Both are equally important. Excellent legal submissions cannot remove URLs that have never been identified, whilst identifying every relevant URL achieves little if the legal submissions themselves fail to persuade.
“The Right to be Forgotten is not just a legal exercise; it is a combination of legal analysis and technical investigation. Artificial intelligence has materially improved one of those disciplines, but it has not replaced the other.”
Preparing the legal submissions is only one part of the process. The second, and often much larger, element is technical investigation, including the identification of every relevant URL.
This frequently involves locating duplicate publications, archived copies, syndicated articles, regional editions, AMP pages, PDF versions, image results, cached pages, mobile variants and numerous other URLs that do not appear during an initial, single Google search.
Search results are not fixed or immovable. The results shown, and their order, change as content is added, removed, republished and re-indexed, and as search engines reassess which results are most relevant to the search term or most authoritative.
This means that the adverse media visible in name-based searches today may be only the tip of the iceberg. Google now limits users to the first ten pages of search results for most queries, whereas historically users could browse much further into the index. Duplicate publications, archived copies and URL variants can surface later as the index reorganises.
This is why identifying the first set of visible URLs is rarely the end of the work. Ongoing monitoring is needed to locate newly visible results and make proactive further submissions before they become prominent. At Internet Erasure we refer to this continuing process as “Sweep and Submit“.
Removing 90% of the adverse media and leaving 10% behind is like swimming 90% of the way across a river and then stopping. In the case of the Right to be Forgotten, one negative result still showing in search results can be almost as damaging as ten. Anyone who finds the remaining article still reads the same story and reaches the same conclusions.
Artificial intelligence is good at analysing information that has already been presented to it. The difficult part is often ensuring that all of the relevant material has first been identified.
With recent improvements, AI is becoming a more useful and valuable tool for research and drafting, like a calculator but for words. It can explain legislation, summarise judgments, improve drafting and identify relevant legal principles, but its output should be checked carefully. Right to be Forgotten cases turn on specific facts, applicable law and supporting evidence, and AI-generated advice or submissions may be incomplete or inaccurate if not independently verified.
As search engines increasingly deploy artificial intelligence across their services, and potentially within the initial assessment of submissions, applicants should assume that clear, well-structured and properly evidenced submissions are becoming more important than ever.
AI can only analyse the material presented to it.
Additionally, AI cannot currently conduct many of the practical investigative tasks that underpin successful Right to be Forgotten work. General-purpose AI systems will not ordinarily be able to sign into social-media accounts, inspect closed groups, access subscriber-only newspaper archives.
Search engines may also continue displaying stale titles or snippets for many years after the actual page has changed. These “cache” links to snippets still stored on Google servers are invisible to AI systems and require manual discovery and checking of the live page against the search result and, where appropriate, a separate outdated-content or cache-removal process.
The investigation must also consider name variations, close misspellings, contextual search terms, search predictions and autocomplete suggestions, all of which can unearth previously unseen results.
“Artificial intelligence and LLMs can analyse legislation, case law and evidence with remarkable speed. They cannot, however, replicate the investigative process of knowing where to look, what to search for, and which additional URLs, variants and publications are likely to surface in name-based searches.”
These limitations are because there is a big difference between analysing information and obtaining it.
Professional judgement also remains important. The Right to be Forgotten is not a mechanical process in which the same facts will always produce the same result. It is a discretionary balancing exercise in which reasonable decision-makers may legitimately reach different conclusions from facts which appear almost the same.
Knowing which arguments to advance, which evidence to prioritise, when to appeal, and when to pursue alternative legal routes often depends upon practical experience rather than simply knowledge of the legislation.
A refusal may require revised evidence, a differently framed submission, further arguments, escalation as a formal complaint or notice to the search engine’s Data Protection Officer (DPO), or, where appropriate, appeals to the relevant data protection regulator.
Theoretically, you can write your own will, complete your own tax return, prepare your own Land Registry plans, represent yourself in court or repair your own car.
That does not mean that DIY (even when supported by AI) is always the best course of action.
In fact, most of us know stories where taking the DIY path proved disastrous (including when I put up shelves at an angle and my daughter’s ornaments slowly rolled off them).
People choose professional representation for a number of reasons. Often to save them time, to know that the successful result is more likely, or certain, and particularly in the case of Right to be Forgotten applications, to avoid revisiting difficult or deeply personal events themselves.
Since February 2022, Internet Erasure has acted in more than 1,500 Right to be Forgotten matters. That experience has provided valuable insight into what separates successful applications from unsuccessful ones.
“The value of professional representation does not lie simply in the depth of legal knowledge. It lies in the ability to combine legal analysis, technical investigation and practical experience to achieve the best possible outcomes for clients.”
The capability scorecard below compares self-representation (DIY submissions), AI-assisted self-representation and specialist representation across ten key capabilities involved in completely clearing search engine results pages of adverse media using the Right to be Forgotten under Article 17 UK GDPR.
With time and determination, simple applications can be made without professional help, or with AI coaching. But in practise, most cases require specialist legal knowledge, technical investigation and strategic judgement. Without experienced handling, many applications are declined or only achieve limited partial removal.
We can see from Google’s Transparency Reports that almost half of the Right to be Forgotten requests they receive are refused. Even where Google have successfully delisted some URL’s, applicants frequently overlook duplicate publications, archived copies, URL variants or additional search results that continue to appear when their name is searched.
There are also procedural considerations.
Multiple unsuccessful submissions can be classified by search engines as excessive, and some self-represented applicants have been notified that further submissions will not be considered for up to 180 days.
Where a UK case must be escalated to the Information Commissioner’s Office, further delay may follow. The ICO currently states that complaints and appeals are being assigned to case officers within 40 weeks of submission. In most cases, during this waiting period, the affected individual continues to suffer the injustice of adverse media sitting at the top of search results and negatively impacting almost every area of their life.
This makes the preparation and timing of submissions an important strategic consideration, particularly in more complex cases.
Professional representation therefore involves far more than preparing legal submissions. It combines legal review, technical investigation, evidence gathering, strategy based on experience and continued monitoring. This maximises the chances of achieving genuinely clear search results within a reasonable period.
Artificial intelligence can assist with parts of that process, but it cannot yet replace the combination of legal reasoning, investigative work and strategic decision-making that complex cases frequently demand.
Artificial intelligence has changed the way many Right to be Forgotten applications are prepared and has made some steps in the process easier to follow for people. What it has not changed or really helped to achieve is the real objective.
“The objective of the Right to be Forgotten is not just to obtain an undertaking of removal of a single set of specific web links which were showing in search results. Instead, the objective is to be able to search the data subject’s name and see no adverse media in search engine results pages. It is only when search results are clear that the Right to be Forgotten has been achieved.”
Internet Erasure Ltd welcomes referrals from a wide range of agencies and law firms acting on behalf of clients. We particularly understand the importance of clearing online search results for legal professionals, as well as individuals facing charges or potential further publicity.
If you would like to learn more about removing content from Google under the Right to be Forgotten, or explore our other related services, please contact us via WhatsApp, email, or telephone through www.interneterasure.co.uk [2].