In the age of Google’s EEAT criteria (Expertise, Experience, Authority, and Trustworthiness), brands and SEOs alike are scrambling to retrofit their content to look, feel, and smell like it was authored by a qualified human. This has become a tired checklist: a stock photo of a smiling expert, a generic bio, a vague nod to credentials, and some bloated prose designed to look 'insightful'.
And yet, most of it is theatre. We've created a cottage industry around fabricating credibility, not earning it.
So what happens when the performance is no longer necessary? What if we could synthesise authority; not as a cheap imitation, but as a legitimate, trustworthy alternative? We’re already halfway there, after all. Much of today’s ‘expert’ content is ghost-written, templated, or heavily optimised for algorithms. That doesn’t scream ‘trustworthy’.
What if AI-created authority - authored not by a named human expert, but generated by a system trained on the sum of human knowledge - becomes not only viable, but preferable?
The problem with human authority
Brands have a fundamental problem with EEAT. Most don't have credible experts in-house. Or if they do, those individuals:
Don’t want to be public-facing,
Aren’t skilled writers,
Pose legal and reputational risks if they leave or get something wrong,
Often lack capacity, training, or support to contribute effectively,
Are disconnected from (or in a different silo to) the product/service, marketing, and/or publishing workflows and content strategy.
And even when brands do publish content under a real name, it's rarely the authentic voice or perspective of the named author. More often, it's a post written by an SEO agency, or an internal content team, attached to an individual for credibility.
So companies fudge it. They invent personas. They buy stock photos. They commission ghost-writers to produce content under inflated bylines. And while not all authors are fake, some are - and many more are figureheads for content they didn’t actually write.
This is not a minor issue of resourcing - it’s a structural failing. Most businesses simply cannot (or will not) scale credible human authority. And they know it.
If we’re already this comfortable with the illusion, why not go further? Why not embrace a model where the source of expertise isn't a human at all, but a system - one trained, tested, and tuned to deliver accurate, helpful, consistent information?
Redefining the E-E-A-T criteria
Google originally advised that content should be written by humans. Today, that guideline is more implied than explicit - but its influence lingers. Reading between the lines of the EEAT framework, it’s easy to see how it was designed to prioritise signals of human authorship, in a sea of machine-generated sameness.
But what if that assumption was a category error? What if synthetic systems could meet – or exceed – each of these criteria?
Expertise has traditionally meant subject matter knowledge. But AI doesn’t just read more than any human ever could; it remembers everything, finds patterns, and evolves with new information. And crucially, it encodes not just facts, but human preferences, biases, and priorities - surfacing insight we may not have noticed ourselves.
Experience has traditionally implied first-hand engagement. But in reality, most published content is reconstructed from second-hand sources. And synthetic systems can already match (and often exceed) those methods. We'll explore this further shortly.
Authority is often measured by visibility, consistency, and coherence - qualities that synthetic systems can deliver with exceptional fidelity. Trained models can operate within clear standards, scale knowledge dissemination, and maintain a stable tone and quality across massive volumes of content.
Trust is built on reliability. And reliability depends more on outcomes than origins. If a synthetic system is accurate, consistent, and auditable - why shouldn’t we trust it? Ironically, AI systems may even hallucinate less than humans. They don’t misremember, embellish, or improvise under pressure. When grounded and properly validated, synthetic systems can be more consistent and less prone to error than fallible individuals with good intentions
If a machine can consistently demonstrate knowledge, contextual awareness, insight, and objectivity - then it is, by all practical measures, an expert. And if it has been trained in a specific domain, tested for rigour, and optimised for reliability, then why shouldn't it be acknowledged as such?
Interestingly, this isn't a new idea. Early AI models were literally called 'Expert Systems'. The difference now is scale, sophistication, and subtlety.
So what happens when we start naming these systems? When we give them faces, bios, credentials? When a synthetic system trained on aerospace engineering, for example, becomes 'AeroAI', complete with a byline and author page?
This kind of symbolic authorship - giving synthetic systems names, avatars, even personalities - might be the bridge between process and trust. It helps users relate to a voice. It enables attribution. And it allows for continuity, improvement, and accountability over time. In this sense, personifying an expert system isn't deceptive; it's clarifying. It tells us what it knows, what it's for, and who’s responsible for its output.
Does that system not deserve the same recognition we afford a junior marketing analyst turned blog contributor?
Perhaps the future of EEAT isn't about proving a human wrote it; it's about proving that the system behind it is credible, transparent, and consistently right.
Simulated experience
Let’s unpack that last point. Of all the EEAT criteria, “experience” is arguably the most difficult to replicate - and the most misunderstood.
We tend to treat "experience" as sacred; a uniquely human attribute. But in digital content, it's already mostly reconstructed from second-hand data. A junior copywriter summarises customer reviews. A freelancer paraphrases a product manual. A ghostwriter interprets a subject matter expert’s bullet points.
Synthetic systems can do all of this - but faster, better, and with a broader dataset. They can ingest thousands of reviews, analyse sentiment, benchmark performance, and identify patterns. But they can also go further: into simulation.
These systems can operate in synthetic environments - virtual sandboxes where they can model real-world scenarios, run repeatable experiments, and surface insights that no human has the time or scope to uncover. For instance, an AI model could simulate 1,000 loads of laundry using a virtual representation of a washing machine, testing how different cycles, detergents, and fabric types affect results - all in seconds.
This isn’t mimicry. It’s something qualitatively new. These systems don’t just report on experience - they generate it, at scale.
And we already accept this kind of synthetic experience in high-stakes domains. Pilots train in simulators. Surgeons rehearse in VR. In the context of Google's "Your Money or Your Life" (YMYL) guidelines - where EEAT is critical in areas like finance, health, and safety - shouldn't we expect more rigour, not less? If AI can provide greater consistency, deeper testing, and broader perspective, isn’t that *more* trustworthy than a human’s one-time experience?
In this light, the idea of AI lacking "experience" starts to look less like a limitation - and more like a misconception.
Implications for search and content
If synthetic systems can fulfil EEAT criteria more reliably, objectively, and consistently than humans, that has enormous implications:
Google's approach to EEAT has always prioritised outcomes - relevance, clarity, helpfulness. Despite early guidance that content should be human-authored, there's little evidence that the mechanism of authorship matters to them; only the result. As synthetic systems improve, it's likely Google will continue to optimise for output quality over process purity.
The fetishisation of human authorship may decline in favour of transparent, process-driven credibility.
Brands might build institutional authority through AI-driven knowledge pipelines instead of relying on fragile individuals.
If synthetic systems can generate better, safer, more reliable information, then the way we think about authorship, trust, and quality needs to change. Google is unlikely to care how content is produced - only whether it’s helpful, relevant, and aligns with user expectations. As long as synthetic content performs well, it's reasonable to expect Google will continue surfacing it, regardless of whether it was written by a person, a system, or a team of both. And for brands, the stakes are even higher. Those who adopt synthetic authority early could build knowledge systems that outperform even the best editorial teams.
If synthetic authority delivers better outcomes, then clinging to the performance of human authorship becomes a liability, and not a virtue. The brands that embrace synthetic pipelines - where content is the output of a calibrated, auditable system - will dominate visibility and trust at scale.
The question isn’t just whether we still want human authorship. It’s whether we can afford it.
Content for machines, not people
Increasingly, content isn’t created just for humans - it’s created for systems. With the rise of zero-click search, AI-generated summaries, and answer engines, much of your content will never be seen directly. It will be parsed, summarised, and repackaged elsewhere.
In this world, content becomes a structured expression of inventory - your products, your processes, your perspectives. And synthetic systems are far better equipped to produce this kind of structured, semantically rich, machine-readable content at scale.
They can work to schema. They can write for APIs. They can optimise for vector embeddings and large language model inputs. They can be consistent, auditable, and exhaustive in ways that human teams simply can’t match.
The goal isn’t just to inform readers anymore. It’s to feed machines - clearly, consistently, and at scale. And the brands that embrace synthetic authorship for this layer of content will build the most durable foundations in the new search ecosystem.
The ethics and optics of the synthetic expert
We're already normalising synthetic authority. Every time we trust Google's featured snippets, AI-generated overviews, or ChatGPT's summarised explanations, we're deferring to machine-generated expertise. We don't ask who wrote it - we judge it by whether it works. If that trust is already in place, the discomfort with synthetic authorship isn't really about quality; it’s about optics.
This shift will be controversial. Some will argue that synthetic authority undermines trust, objectivity, and accountability. Others will point to the uncanny valley effect - the discomfort we feel when machines mimic humans too closely - and suggest that society may never be fully comfortable with synthetic personas.
There are valid concerns here. What happens when a synthetic expert gets something wrong? Who takes responsibility? What if a system is optimised to persuade rather than inform? What if bias is amplified rather than neutralised?
But today’s status quo isn’t exactly virtuous either. We already tolerate ghostwriting, stock authorship, and algorithmically optimised fluff. Many articles are commissioned purely to rank, created on a conveyor belt by outsourced writers with no disclosure, context, or connection to the reader.
Synthetic authority isn’t about deception - it’s about moving beyond that performance to something more rigorous, auditable, and transparent.
And transparency is the real issue. Should readers be informed when content is machine-generated? Possibly. But should they also be told when it was ghost-written? When it was crafted for SEO, not for users? When it’s technically correct but experientially hollow? Transparency cuts both ways.
There are already emerging legal frameworks that require disclosure of AI-generated content in specific contexts. But the more important question isn’t whether AI wrote it. It’s whether the method of authorship - human or synthetic - is clear, traceable, and held to a standard.
Synthetic authorship doesn’t lower the bar. It raises the expectation that we explain how things are made, and why.
Toward a post-human web of trust
We may be entering a phase where the most credible, helpful, and consistent information online comes from non-human systems. Authority is being unbundled from identity. Expertise is becoming a property of process, not personality.
Synthetic authority isn’t about replacing people. It’s about building a new model of trust - one based on consistency, transparency, and performance. Done well, it can be more rigorous than legacy authorship, more honest than ghost-writing, and more scalable than relying on a handful of named experts.
Before you can operationalise synthetic authority, you need a clear foundation:
Identify your knowledge domains. What topics does your organisation need to own authoritatively?
Start curating and categorising knowledge. That means source material, internal documentation, first-party data, and structured feedback loops.
Design and train your expert systems. Whether through fine-tuned LLMs or structured workflows, begin to systematise how your expertise is captured, reviewed, and published.
Establish transparent authorship protocols. Declare when a system is the author. Show its training data. Explain how it’s maintained.
The challenge now isn’t just whether we allow synthetic experts. It’s whether we treat them seriously - whether we build them, govern them, and demand more from them than we ever did from their human counterparts.
The brands that embrace synthetic authority (openly, ethically, and strategically) will redefine how trust and knowledge work online. And those who cling to nostalgic notions of human-only authorship may find themselves both outpaced and outclassed.
The age of synthetic authority isn’t coming. It’s already here. The question is whether we’re ready to trust it – and whether we’re prepared for a world where the most authoritative content isn’t produced by individuals, but by processes.
Article by
Jono Alderson
Jono Alderson is an award-winning digital strategist and SEO consultant, known for his expertise in technical SEO, performance, structured data, and WordPress. He helps brands grow, win markets, and prepare for the future—code, content, and all.
stay in the loop