Generative AI has, in the space of about 30 months, made it dramatically easier for someone to look qualified for a job and dramatically harder for an employer to tell whether they actually are. The asymmetry runs in one direction. Candidate-side tools have flooded the market with frictionless CVs, polished cover letters, and AI-burnished portfolios. Meanwhile, recruiter-side tools that traditionally separate signal from noise—screening, interviews, references—have not improved at remotely the same pace. The result, by mid-2026, is a hiring market in which the cheapest input has scaled fastest and the most expensive one, the recruiter’s time, has been overwhelmed. Ethos, a London-based AI startup founded by alumni of Google DeepMind and McKinsey, has decided that this is a fundable problem. On Wednesday morning, the company announced a $22.75m Series A funding round led by Andreessen Horowitz, with participation from General Catalyst, XTX, and Evantic. It is one of the larger Series A rounds for a UK AI startup this year.
The five core ai hiring flaws Ethos targets
Before exploring what Ethos actually builds, it helps to understand the five specific ai hiring flaws the company aims to fix. Each one represents a breakdown in trust, verification, or efficiency that has worsened since generative AI became mainstream. Ethos does not try to solve every problem in recruitment. It focuses on these five.
Flaw one: Credential inflation without verification
A resume today can be generated in seconds. A chatbot can produce a cover letter that sounds authentic. A portfolio can be padded with projects that never happened. The hiring market now suffers from what economists call an asymmetric information problem. Candidates can inflate their credentials with almost zero cost. Employers, on the other hand, must spend hours or days to verify each claim. This imbalance means that dishonest or exaggerated applications crowd out honest ones. Ethos addresses this by building verified expert profiles that go beyond what a candidate chooses to disclose.
Flaw two: Static documents cannot capture tacit knowledge
A CV is a list of titles and dates. It tells you where someone worked and for how long. It does not tell you how they think, how they solve problems, or what nuanced expertise they possess. Much of professional knowledge is tacit. It lives in the way a surgeon decides where to cut, how a lawyer constructs an argument, or how a plumber diagnoses a leak. These are not things a resume can convey. Ethos uses an AI voice agent to interview experts, surfacing the texture of their professional knowledge in a way a static document cannot.
Flaw three: Recruiter time is the bottleneck
Before generative AI, a recruiter could reasonably screen fifty applicants for a single role. Today, that same recruiter might face five hundred applicants. The cheap input—applications—has scaled dramatically. The expensive input—human judgment—has not. This creates a bottleneck. Recruiters must either spend more time per role or lower their screening standards. Neither option is sustainable. Ethos automates the matching process, using AI to pair verified experts with opportunities at a scale no human team could match.
Flaw four: Traditional expert networks are slow and expensive
Companies like GLG and Guidepoint have built valuable businesses by curating rosters of consultants and domain specialists. But those rosters are built by humans. They take years to develop)Skip and rely on manual vetting. They are expensive to maintain and slow to scale. When a frontier model lab needs a verified cardiologist for training data within a week, a traditional expert network cannot deliver. Ethos builds its profiles with AI, then matches them autonomously, compressing a process that used to take weeks into hours.
Flaw five: AI training data lacks verified domain expertise
Frontier model labs need high-quality, domain-specific training data. General-purpose web scrapes are insufficient for fields like finance, medicine, law, and advanced engineering. Yet the current market for expert training data is fragmented and opaque. A lab might pay a platform for access to experts, only to discover that those experts have inflated their credentials or lack real-world experience. Ethos positions itself as a route through which labs can access verified domain experts at scale, closing a gap that threatens the reliability of AI systems themselves.
What Ethos actually builds
Ethos is, in plain terms, an AI-driven expert network. Where traditional expert networks rely on human curators, Ethos uses AI to do the curation. The mechanism, as described on the company’s product page, is two-pronged. First, an Ethos voice agent conducts an extended interview with each expert. This conversation surfaces the texture of their professional knowledge in a way a static CV cannot. Second, Ethos’s AI ingests the expert’s existing portfolio of work, including academic papers, code repositories, blog posts, podcast appearances, and conference talks. The combined profile is then matched, autonomously, against opportunities flowing in from the platform’s customer base.
The voice agent advantage
The voice agent is not a simple chatbot. It asks follow-up questions. It probes for depth. It can detect when an expert is glossing over a topic and push for more detail. This process creates a rich, multidimensional profile that captures not just what someone knows, but how they think. For an employer trying to distinguish between two candidates with similar resumes, this kind of depth is invaluable. It is the difference between knowing someone’s job title and understanding their judgment.
Ingesting the full digital footprint
Most hiring platforms rely on self-reported data. Ethos goes further. By ingesting an expert’s actual work product—papers, code, talks, podcasts—the platform builds a picture that cannot be faked. A candidate can claim expertise in machine learning, but if their GitHub repositories are empty, the profile will reflect that. This approach shifts the burden of verification from the employer to the platform, which is exactly where it should be in an AI-mediated hiring market.
Autonomous matching at scale
Once a profile is built, Ethos matches it against incoming opportunities automatically. A consulting engagement might require a former CFO with experience in healthcare. An AI data-labelling project might need a licensed electrician with knowledge of commercial wiring codes. The system handles both, using the same underlying model. This scale is what traditional expert networks cannot achieve. A human curator can manage a few hundred relationships. Ethos manages hundreds of thousands.
The funding story and what it signals
The $22.75m Series A round led by Andreessen Horowitz is one of the larger rounds for a UK AI startup this year. General Catalyst, which led the seed round in 2024, participated again. XTX and Evantic also joined. The size of the cheque tells you something about how seriously a16z is taking the labour-market dimension of AI’s commercial impact. Hiring is not a niche problem. It is a multi-billion-dollar market that AI has visibly degraded. Ethos is betting that the solution is not better resumes, but better verification.
Why Andreessen Horowitz is interested
Andreessen Horowitz has a track record of investing in infrastructure that sits between AI models and real-world applications. Ethos fits that pattern. The company is not building a better job board. It is building a verification layer for the entire labour market. If that layer works, it becomes a critical piece of infrastructure for every company that hires knowledge workers. That is a large addressable market, and a16z wants to own it.
The timing of the round
The round comes at a moment when hiring is becoming the part of the labour market AI has most visibly degraded. Candidate-side tools have flooded the market with frictionless applications. Recruiter-side tools have not kept pace. Ethos is positioning itself as the counterweight, a system that restores trust to a process that has lost it. The timing is not accidental. The problem has become acute enough that companies are willing to pay for a solution.
Traction and growth metrics
The traction figures in the announcement are the kind that, if accurate, justify the Series A size. The company says that more than 5,000 experts join the platform each week across accounting, banking, consulting, law, technology, and healthcare. Skilled tradespeople including electricians and plumbers also join. The cross-collar reach is unusual for an expert network and consistent with Ethos’s broader pitch: the unit of value is verified expertise, regardless of the credentialing path that produced it.
Earnings on the platform
The average expert on Ethos earns £4,500 per month in additional income through the platform. The top 10% make more than £7,000 per month. Since January, the number of experts earning income through Ethos has grown six-fold. Per-hour rates range from $105 to $225, materially higher than standard AI-training pay tiers. These figures suggest that the platform is attracting high-quality experts who command premium rates.
Supply and demand dynamics
Whether those figures hold under scrutiny will depend on the durability of the underlying customer demand. Ethos matches experts to consulting engagements, expert calls, market research surveys, AI data-labelling projects, and full-time roles. The AI-data line is structurally important. Frontier model labs need verified domain experts for training data, and Ethos offers a pipeline that traditional networks cannot match. If demand from those labs continues to grow, the platform’s economics will strengthen.
How Ethos compares to traditional expert networks
Traditional expert networks like GLG and Guidepoint have spent decades building human-curated rosters. Those rosters are valuable, but they are also expensive and slow to scale. Ethos takes a different approach. Instead of feeding existing expert networks into AI products, it builds the expert profiles themselves with AI, then matches them to opportunities at the scale only a model-based system can manage.
You may also enjoy reading: 7 Ways Nio Onvo L80 Undercuts Tesla in China.
The GLG and Guidepoint comparison
GLG and Guidepoint have themselves now become data partners inside frontier models like Claude Opus 4.7. That is a sign that traditional expert networks have value, but it also highlights their limitations. They are being used as training data, not as active matching systems. Ethos is building a counter-pattern. It uses AI to curate profiles and then matches them in real time. This is a fundamentally different business model, one that is faster, cheaper, and more scalable.
The scale advantage
A human curator at a traditional expert network might manage a few hundred relationships. Ethos manages hundreds of thousands. The platform adds 5,000 experts per week. That scale is possible only because the curation process is automated. The voice agent interviews candidates. The AI ingests their portfolios. The matching engine pairs them with opportunities. Humans are involved only when the system flags an edge case. This is the difference between a craft business and a platform business.
The AI data-labelling opportunity
Frontier model labs need high-quality, domain-specific training data. General-purpose web scrapes are insufficient for fields like finance, medicine, law, and advanced engineering. Ethos has positioned itself as a route through which those labs can access verified domain experts at scale. This is a structurally important revenue line for the company, and one that traditional expert networks struggle to serve.
Why labs need verified experts
When a frontier model lab trains a model on legal reasoning, it needs data from real lawyers. When it trains a model on medical diagnosis, it needs data from real doctors. Using unverified sources introduces risk. A model trained on bad data will produce bad outputs. Ethos offers a pipeline to verified experts, reducing that risk. For the labs, this is not a nice-to-have. It is a requirement for building reliable AI systems.
The pricing advantage
Per-hour rates on Ethos range from $105 to $225. That is higher than standard AI-training pay tiers, which often fall below $50 per hour. But for a frontier model lab, paying $200 per hour for a verified expert is a bargain compared to the cost of training a model on bad data. The premium is justified by the verification. Ethos is betting that labs will pay more for certainty.
The cross-collar reach
Ethos does not limit itself to white-collar professionals. Skilled tradespeople including electricians and plumbers also join the platform. This cross-collar reach is unusual for an expert network)Skip but it is consistent with Ethos’s broader pitch. The unit of value is verified expertise, not the credentialing path that produced it. A plumber with thirty years of experience knows things that no AI model can replicate. Ethos captures that knowledge and makes it available to customers.
Why tradespeople matter
The inclusion of skilled tradespeople reflects a recognition that expertise is not limited to university degrees. A master electrician who has wired hundreds of commercial buildings has tacit knowledge that is valuable in AI training data, market research, and consulting engagements. Traditional expert networks have largely ignored this population. Ethos sees it as an opportunity.
The verification challenge for trades
Verifying a tradesperson’s expertise is different from verifying a consultant’s. There are fewer public records. There are no academic papers or conference talks. Ethos addresses this through the voice agent interview, which can surface the depth of practical knowledge. A plumber might not have a GitHub repository, but they can describe in detail how they diagnose a hidden leak. That kind of knowledge is valuable, and Ethos captures it.
The future of hiring in an AI-mediated market
If Ethos succeeds, it will change how companies think about hiring. The platform shifts the burden of verification from the employer to the platform. It replaces static documents with dynamic, verified profiles. It automates matching at a scale no human team can match. These changes address the five core ai hiring flaws that have emerged since generative AI became mainstream.
The broader implications
The hiring market is not the only market AI has degraded. But it is one of the most visible. Every company that hires knowledge workers faces the same problem: too many applications, too little signal. Ethos offers a way out. It does not solve every problem in recruitment. But it solves the verification problem, and that is the one that matters most right now.
What comes next
Ethos will need to prove that its verification process is robust enough to withstand adversarial attacks. It will need to show that its matching algorithm is accurate and fair. It will need to demonstrate that its economics work at scale. The Series A funding gives it the runway to do all of these things. Whether those figures hold under scrutiny will depend on the durability of the underlying customer demand. But for now, Ethos has a clear thesis, a strong team, and a large cheque. That is a good place to start.
