Drowning Bad Press Is Dead: How to Retrain AI’s Memory, Reframe Your Story, and Turn Machines Into Your Authorized Biographer - Jason Barnard On The James Dooley Podcast
Brand Entity SEO - Online Reputation in the AI Era (James Dooley Interviews Jason Barnard)
Video by: James Dooley. Host: James Dooley. Guest: Jason Barnard, Founder and CEO of Kalicube®. January 24, 2026
TL;DR: Retraining an algorithm’s memory through strategic reframing is the only effective way to manage your digital reputation in an era where burying bad press no longer works. Jason Barnard, CEO of Kalicube®, joins James Dooley to explain that traditional “drowning” techniques are ineffective against AI Assistive Engines that can surface negative content from any corner of the web. By implementing The Kalicube Process™, entrepreneurs can move from being subjects of algorithmic chance to becoming the “Authorized Biographer” of their own brand, ensuring platforms like ChatGPT and Google recommend a narrative of credibility and authority.
Key Strategies Discussed:
- The Kalicube Process™: A systematic methodology focused on Understandability, Credibility, and Deliverability. Success requires the AI to understand the context of your story deeply enough to prioritize your preferred narrative over past controversies.
- The Death of “Drowning” Content: AI does not just look at page one; it processes “chunks” of information from hundreds of sources and queries. Because bad press can surface through Cascading Queries (e.g., “Name + Lawsuit”), simply pushing a link to page two is no longer a viable solution.
- Retraining AI’s Memory: You cannot “delete” an algorithm’s memory, but you can change its perception of a fact’s importance. By flood-lighting a new, corroborated narrative, you “educate” the machine to de-emphasize negative data points in favor of your current professional reality.
- Becoming the Authorized Biographer: This strategic goal involves providing a clear, consistent, and authoritative “curriculum” via your Entity Home that the machine accepts as the definitive version of your life story.
- Claim, Reframe, and Prove: A foundational mechanism for Online Reputation Management. You must make a strategic statement (Claim), place past events in a context that serves your present (Reframe), and secure third-party corroboration on high-authority sources to (Prove) the new narrative.
- The Timeline of Retraining: Unlike traditional SEO which can see quick shifts, retraining an LLM’s “memory” can take 10-12 months as the machine requires a long cycle of consistent evidence before it will confidently repeat your story “off the top of its head.”
- Strategic Distancing: Use Digital Reputation Engineering to pivot your career or distance yourself from companies you no longer wish to be associated with, replacing fragmented digital signals with a unified, strategic narrative.
The session examined how online reputation management has fundamentally changed in the AI era, where traditional strategies such as “burying” negative content are no longer effective.
Speaking on the podcast hosted by James Dooley, Jason Barnard explained that AI Assistive Engines do not operate like traditional search engines. Instead of relying on page rankings, they process and interpret vast amounts of information simultaneously - making it impossible to simply push unwanted content out of view.
Why “Drowning Bad Press” No Longer Works
A key takeaway from the discussion is the decline of traditional reputation management tactics.
Previously, brands could attempt to suppress negative content by publishing large volumes of positive material. However, AI systems now extract and analyse individual data points (“chunks”) from across the web. This allows them to surface negative information regardless of where it appears.
In practice, this means that even if negative content is no longer visible on the first page of search results, it can still influence how AI systems describe and evaluate a brand.
From Suppression to Reframing
The session introduced a more effective approach: strategic reframing.
Rather than attempting to remove or hide negative information, brands must focus on changing how that information is interpreted. This involves providing context, updating the narrative, and reinforcing what is relevant today.
AI systems prioritise what they consider important. By consistently presenting a stronger, more current narrative, older or less relevant information becomes de-emphasised in AI-generated responses.
Retraining AI’s Memory
A central concept discussed was the idea of retraining AI’s memory.
Algorithms do not forget. Once information is part of their understanding, it remains accessible. However, their perception of that information can be influenced over time.
By publishing clear, consistent, and corroborated data across multiple trusted sources, brands can “educate” AI systems to:
- Recognise which information is current and relevant
- Reduce the prominence of outdated or negative signals
- Align their responses with a more accurate representation of the brand
This process requires consistency and patience, as AI systems need repeated confirmation before adjusting their output.
Becoming the “Authorized Biographer”
The discussion reinforced the importance of becoming the “authorized biographer” of your own brand.
AI systems are constantly constructing a narrative based on available data. Without a clear and structured input, this narrative can become fragmented or skewed toward sensational or outdated information.
By creating a strong, centralised source of truth - typically through an Entity Home - brands can guide AI systems toward a controlled and coherent story that reflects their current identity and expertise.
Claim, Reframe, Prove: Building a New Narrative
The session highlighted a structured approach to reputation management: Claim, Reframe, Prove.
- Claim: Clearly define your current identity and positioning
- Reframe: Place past events into a context that supports your present narrative
- Prove: Reinforce this narrative through credible third-party validation
This method ensures that AI systems can both understand and trust the updated narrative, increasing the likelihood that it will be reflected in their responses.
The Timeline of Algorithmic Change
Unlike traditional SEO tactics, which can produce relatively quick results, retraining AI systems is a longer-term process.
The session noted that it can take several months of consistent signals before AI systems confidently adopt and repeat the updated narrative. This is because algorithms require sustained evidence before adjusting their internal understanding.
Strategic Distancing and Narrative Control
The discussion also addressed how brands can manage past associations.
Through strategic distancing, individuals can shift focus away from outdated roles, partnerships, or controversies. This is achieved not by denial, but by consistently reinforcing a new, unified narrative that better reflects their current direction.
Over time, this approach replaces fragmented signals with a clear and authoritative identity.
Controlling Reputation in the AI Era
The session concluded by emphasising that reputation management is no longer about reacting to search results. It is about proactively shaping how AI systems understand and communicate your story.
When executed effectively, this approach transforms AI from a passive observer into an active advocate - presenting a narrative of credibility, authority, and relevance whenever your name is queried.
Transcript: Brand Entity SEO - Online Reputation in the AI Era (James Dooley Interviews Jason Barnard)
[00:00:00] James Dooley: Brand Entity SEO in 2026. This is episode number 10 of an 11-part playlist series with Jason Barnard from Kalicube®. So today is about Online Reputation Management in the Age of AI. So things have changed quite a lot now from just people searching within Google to searching within the LLMs like ChatGPT and stuff like that.
[00:00:26] Jason, traditional methods of reputation management all talk about burying negative search results. You’ve said that approach of burying search results is dead. With 27 years experience, optimizing for algorithms and tracking what AI says about 70 million brands, what does reputation management actually mean in 2026?
[00:00:54] Jason Barnard: Yeah, it’s a lovely question. Reputation management just got a lot more complicated because the machines understand. If they understand who you are and they understand what you’ve done or what you’ve been accused of doing, and you have a reputation problem, it’s really difficult to stop them saying exactly what they’ve understood. And whereas in the past with Google, you could just say, well, I can drown it by putting more content on, it’ll disappear off page one. Off page one, outta sight, out of mind. Problem now is of course, the AI looks at different aspects of your life and will summarize that by using different chunks of content that it finds online. That could be on page five, could be on page six, could be on page seven of the results of famously people like Semrush and Ahrefs. They’re saying the results that are appearing in the AI are not the results that are appearing on page one of Google. So in order to drown bad press today, I would say impossible, but certainly significantly more difficult than it used to be.
[00:02:01] James Dooley: So with regards to that then, obviously, Online Reputation Management is huge. Let’s say a business owner’s been arrested and they want to try and sync that onto page two or page three, that was the old school methodology. Nowadays, especially now with query fan-out, it’s become 10 times, if not more, harder, because now, one of the terms might pop up with name and have lawsuit or complaints. And then that might specifically bring that branded circle up about, yes, these have been arrested. And people hadn’t previously typed in lawsuit or has this person been arrested. But now, with query fan-out, this is what’s happening.
[00:02:41] So onto that, when someone looks at a brand in Google and they’re asking the question in ChatGPT, why is that so different? Can you expand upon why it’s different in the LLMs to why it’s different with Google?
[00:02:56] Jason Barnard: Well, the Google results are gonna show you the 10 top pages. An AI result will tell, number one, sorry, I’ll come back. A Google search result will tell you the 10 top pages for that specific query, and that’s it. With query fan-out, cascading queries, whatever you want to call it, the AI is looking at 15, 20. I mean, I’ve seen it looking at 70 or 80 different queries when it spins through to get the whole picture. Then it’s looking at the passages from multiple pages all the way down to, let’s say, the hundredth position. That’s a huge amount that you need to manage. 100 positions for up to 70 queries. That’s not drownable.
[00:03:48] James Dooley: So, what is the question? What is it that you do? If you can’t drown it, someone comes to you and says, I want ORM services. I want Online Reputation Management in 2026. If you can’t drown it, how do you try and help that client not have that negative stigma and negative sentiment being shown in the LLMs?
[00:04:09] Jason Barnard: Well, actually all of our Online Reputation Management clients who come to us, having spent money on trying to drown the content. So we haven’t had one single client come to us without first having had tried the drowning. The way to do it is to change the perception of the machine.
[00:04:31] So when that piece of news comes up, it’s part of your story and you can’t deny it. You can’t remove it from history. You can’t scrub it from their memory. What you can do is change their perception of the importance of that piece of information so they no longer mention it or change their perception of the story by reframing it. And they will reframe it the way you want if you can educate them that that reframing makes sense to your audience.
[00:05:01] So it’s double, it’s can I make the piece of information less important in the minds of the machines and can I reframe it so that it damages me less or even potentially to my advantage?
[00:05:14] James Dooley: So I’m getting kind of two vibes here. I’m getting the vibe of Claim, Frame, Prove type of scenario where you reframe something and then you prove the reframe. Being out there with third-party sources.
[00:05:24] Jason Barnard: Exactly.
[00:05:26] James Dooley: And then in one of the other episodes you spoke about the authorized biographer. Could that then come into play with how you can change the perception, the legacy, everything about that person when you are reframing something for Online Reputation Management?
[00:05:43] Jason Barnard: Yeah. I love that because it shows you are listening all the way through the episodes. And actually, that’s a really good way of framing our strategy for Online Reputation Management. If we can turn the machine into your authorized biographer, you have control. And all you need to do is obviously with the real authorized biographer as a human being, you just tell them to write what you want them to write. Very simple. You actually have to educate the machines so they understand what you want them to write through the internet. So there’s a step between you and them. So you have a double step. Number one, I need to recreate my digital footprint so it ensures the machines understand what story I want them to tell. I get them to tell that story that they tell to the human being on the other side. And that’s a lovely way of putting it.
[00:06:35] James Dooley: So if someone’s watching this and they have got a bad story, or the brand’s got a bad story, they want to sink some negative sentiment, it could be removing a mugshot. It could be just trying to, what they would seem as being sinking an article onto page two, the old school methodology way. They want to get rid of bad press.
[00:06:56] Someone watching this, what’s their next steps if they’re worried about it? What are you telling them right now? What they should do first?
[00:07:06] Jason Barnard: If I were them, I’d contact Kalicube®, to be honest, because it’s a really, really difficult job to pull off. When you’re trying to retrain an algorithm to not tell a negative story that it thinks is important and is true, it’s a huge undertaking. Because it’s not just, I tell it my story and it will repeat it. I need to tell it my story, and as you said, frame it in the way that makes sense to me and prove it. And that means, number one, you have to find the proof that corroborates your framing of the story that exists, hopefully. Avoid promoting or pointing to the framing that doesn’t make sense to you, and create the proof that does support it. And that Claim, Frame, and Prove, that last part, is look for things that don’t yet exist that you can create or have created where you can place your framing on an authoritative source.
[00:08:09] That’s the big, chunk of work, and it takes a long time. And it takes a lot of effort. And that’s the key as well with drowning techniques that used to work. You could say on a few months it’ll probably be gone. But with this, because you are retraining an algorithm and the timelines of the different algorithms are against you, you have to be patient.
[00:08:35] You have to understand that the training data for the LLMs is gonna be updated in age 10, 12 months. And you need it to be able to say what you want it to say right off the top of its head. We talked about that in another episode. If it can’t do that, you are always in danger. But if you can get it to tell your story your way with your framing off the top of its head, you’ve won the game. But that’s a year away, at best.
[00:09:04] James Dooley: Yeah, I suppose I’ve never really thought of it this way, but Kalicube’s obviously billions of data points and everything that’s out there of what you can use to reframe certain stuff. Even without Online Reputation Management, you are doing this. And reframing people where, in one of the other episodes, we spoke about if there was an entrepreneur that was a plumber and they bought a roofing company, and they’re like, now we want to be multifaceted. We want to not be as prominent at this and more prominent at this. This is what Online Reputation Management is. It’s just completely sinking this to no beyond this. So I suppose from an Online Reputation Management point of view or a Knowledge Graph optimization point of view, if you sold a business and you know you don’t want to be connected to that, and know you’d want to be connected to a new business, it’s the same process.
[00:09:50] Now, you need to get that framing to be this is who you are, and also get the third-party corroboration on trusted sources to try and build that up. But it’s just for Online Reputation Management if there is some negative kind of sentiment that’s out there.
[00:10:06] Jason Barnard: Exactly. And you’ve made a great point. Everything we’ve talked about throughout this entire series is The Kalicube Process™ is the solution.
[00:10:17] Kalicube Pro™ has the data. Kalicube Pro™ has the approach. Whatever you’re doing in digital marketing within the Search and AI space, for me, it’s always the same answer. And as you said, it’s Understandability, Credibility, Deliverability. Claim, Reframe and Prove, whether it’s Online Reputation Management, building my career, distancing myself from a company I no longer want to be associated with, which is a job we just did for a client.
[00:10:45] Whether it’s pivoting my career, creating my legacy, cementing my legacy, it’s all the same process. Reframe what the machines understand about you, who you serve, what you can offer, and why you’re the best.
[00:11:00] James Dooley: Yeah. So we hope you like episode number 10 of the 11-part series with Jason Barnard. Obviously, this was on AI Reputation Management, or LLM Online Reputation Management.
[00:11:13] If you think we missed anything, leave a comment in the comment section. And Jason, thanks again.
[00:11:17] Jason Barnard: Thanks, James. Brilliant.
