The Trust Infrastructure Imperative | Global Interview Security Summit 2026 | Opening Keynote [Sanjoe Tom Jose]

Why hiring integrity is no longer an HR problem — it's a business risk that starts at the top.

Key speaker:

• Sanjoe Tom Jose (CEO at Talview)

Transcript

SPEAKER_00 0:00

The biggest threat to IT hiring today isn't talent shortage. It's fraud at scale. According to a joint IBM XFORS and FLAIR report, a coordinated network of 100,000 plus fake IT workers is embedded in real jobs across 40 plus countries, generating over$500 million annually. The engineer who aced your technical interview may not be the one who shows up on day one. Today, IT hiring has become the front line of AI-enabled fraud. Proxy developers clearing coding rooms, AI coaching during interviews, deep fake identities on live panels. The FBI warns about fraudulent IT workers entering companies through compromised hiring. Xperian names deep fake candidates a top fraud threat for 2026. 62% of hiring professionals say candidates are better at faking identities than hiring teams are at spotting them. A fraudulent IT hire can gain access to code bases, infrastructure, and customer data, turning a hiring mistake into a security incident. This is no longer just a hiring challenge, it's a security challenge.

SPEAKER_01 1:24

The trust infrastructure has been the talk of the town for the last few months. With the advancements in AI, everyone is worried about what kind of fraud is happening in the interview process, what kind of threats are filtering through the hiring process and entering your organization undetected and is compromising the quality and the security of your hiring process and the security of the organization. And over the past few months, hiring integrity has become a problem beyond HR. It's a business risk that's starting at the very top. And throughout this summit, we will have conversations from some of the leading experts, some top practitioners, some product technologists who are going to talk about the nature of problems they are facing, the kind of challenges they are seeing on the ground, the solutions they have implemented, the response and success of some of those solutions, the failures or the challenges they still continue to encounter. And also talk about some of the very promising work which Talvi is doing in this particular domain. So let's start with Imperative. Why hiring integrity is no longer an HR problem. The last few months we have all come across stories of organized operators, organized operators from North Korea, from Nigeria, other countries who are trying to infiltrate top employers in the technology space, in the banking and financial services space, in the healthcare space, so that they can steal your data and sell it in the dark web or ask for answer. That is very different from the kind of challenges which HRE is used to facing. It's no longer a hiring quality problem, it's a problem which is posing significant risk to business. And the battlefield has changed. Obviously, AI is the primary reason we are facing this risk. Candidates can apply if they, especially if they're malintent, using ports to hundreds of thousands of job applications. They can file them at one click. And every applicant will have a resume and a covering letter, which is very customized for that particular job, that it's very difficult to distinguish between who is genuine and who is not. You have deep fake technology, so right from images to videos, which are hyper-realistic, that with naked eye you cannot detect which is a true human and who is probably using a deep fake to hide behind the uh image. A lot of organized operators who have entered the space. There was always cheating as service uh companies out there, but they were largely constrained by the number of people who they had on their payroll to take interviews or help candidates take interviews. But now with AI, they are all scaling their operations significantly, which is making it very difficult for us to stay on top of this. And anytime a leak happens, the interview content, the questions, exams that get spread across the globe at a lightning pace that you do not control the narrative. And the threats which are amplified by technologies like crypto, where you know organizers or these some of these uh cheering as service providers can ask for ransom uh in in Bitcoin or some of these uh cryptocurrencies make the problem even more sinister. The scale of the problem is mind-numbling. In terms of numbers, almost 62% hiring professionals who have been recently polled uh have commented that they believe there is fake identities uh being used by candidates in the hiring process. Almost 500 million plus state back fraud, especially from some of the top sponsors of uh state-backed fraud like North Korea, has been operating in IT worker schemes, which is infiltrating many of the organizations we all are part of or are aware of. Uh, by some counts, there are almost 100,000 fake IT workers. Some of them have real humans uh operating from a remote geography, some of them are not even humans, but a fake identity is spun up and maintained with the help of AI agents in the IT industry right now. And on an average, including the direct cost of onboarding provisioning and remediation, almost$33,000 is lost per every fraudland hire. And this is not including the indirect cost if there is a data breach or there's loss of contract and anything else which is associated with these risks. And the traditional interview screening mechanism is failing. You have some of these tools which are available which uh you can use for checking the identity of the candidate, but the candidates know which step you're going to check their identity for, most probably upstream at the application stage or during the background screening stage. You are not continuously checking for who they are throughout the hiring process. You might have the candidate joining in with the recruiter, but when they're meeting the hiring manager for a technical interview, it's somebody else joining. So that lack of cross-interview memory becomes a major challenge. A lot of these uh reports which you get, uh, especially if you're using some of the uh interview proctoring tools out there, but there are these reports have a lot of false positives often and has a lot of different alerts, which leads to an alert fatigue paralysis, which makes it difficult to stay on top of these reports. No cross-intelligence across organizations. There are a lot of uh organized operators who are using the lack of coordination between organizations to infiltrate one organization at a time, or attack an organization in uh China, and then uh let's say attack some them in India or in India. Uh, so and most often, even if even within the same organization, there is not enough data sharing that you are tracking and are able to stay on top of this. And a lot of the tools which are being used today, those tools are reactive and not proactive, which may and makes it very difficult to catch the fraud at the time it's happening, and by the time you detect and you try to act, it's probably too late. So, an interesting analogy is before almost 19 out of the 20 candidates whom you were hiring, they were uh genuine candidates. But in 2026 and beyond, five out of twenty candidates are expected to be a fraudulent higher. So who is checking the tickets before the board? And that's where uh I want to talk about a new paradigm shift which Talview is very proud to launch. We are calling it Assessment Intelligent Operations Platform or AIOP, which is taking a single interview security approach to enterprise-wide hiring risk management. Many of you are already familiar with Talview's patented industry first seven-layer framework for comprehensive interview security, uh, where we tap into different hiring signals to ensure that we are staying on top of what's happening within an interview. But that kind of comprehensive approach uh has some gaps. A lot of data points which you're generating means that you end up with a lot of data overload, which requires a much better aggregation and visualization and real-time intervention. A lot of focus of seven life security is during the interview, how we can protect you when an interview is happening. But the the kind of rich data which we are collecting could actually help you to many times proactively prevent risk with pre-interview risk analysis, as well as also continuously keep track of what's happening post-interview with a lot of data which we are collecting. And at an organization level, what we realized is there is a need for a unified trust view to understand where in the organization, where in the hiring process you are experiencing maximum risk and take control of risk and address risk proactively. AAOP or Assessment Intelligence Operations Platform closes all three, provided by Italy's proprietary ontology, which borrows elite intelligence blueprints from cybersecurity, finance, and defense. And I'm very excited about this new launch. Before we jump into the details of AIOP, wanted to spend a few minutes as a quick refresher for the Talvie's Battle Tested Seven Layer Security Framework. It is the most comprehensive approach to interview security and hiring security today in the market. It has everything from identity verification, ensuring that the candidate is not violating any interview rules using the primary camera, using a secondary camera to understand what's going on in their environment, ensuring that they are not running any of these AI assistance tools like Cluli or final round in their device by ensuring device security, continuously monitoring their responses, whether it's written or oral for any form of uh AI detection using uh tapping into their content feed, looking for any indication of OIS prompts or other forms of OIS-based cheating, including deep fake. And lastly, also ensuring that there are no flags for the candidate from their device, the IP, anything which we have uh previously seen as risk factors when it comes to hiring process. Now, AAOP sits on top of our seminal security framework. It takes all these seven signals, but not just from one interview. It takes signals from across the hiring process, the application process, exams, multiple interviews, onboarding, all of that. Takes that uh into what we call a living ontology incident management engine, which essentially has all the learnings which Talvi has gathered over the years from doing millions of interviews on the platform, and we have prepared a proprietary ontology, which rather than treating each event as an independent flag, which diludes you with uh a lot of data, we actually use an incident management engine to flag real 100% trustable events, which is not just uh looking at what's happening in the interview, but also looking at what's happening in in your organization across uh the borders, across geographies, across business units, functions, and also looking at some of the metadata which we have from across our customers and pooling that with existing non-fraudulent actor metrics, risk vectors, which we have curated from interviews over millions of interviews across multiplayers. And we broadly break this into three different stages in the uh interview process. The first is the pre-interview intelligence, where we are contextually risk profiling each candidate even before they show up. So we look for specific patterns which we have identified as prominent risk vectors in our candidate pool, which includes uh supporting high-risk proxy clusters. So if you are seeing a particular type of IP or a device ID, or if you're seeing uh seeing any IDs or any form of tampering with the uh added proof, which the candidate has submitted as a part of the uh uh application process, or if you're seeing that there are specific requests which are coming in from the candidate which has been associated with uh non-fraudulent actors, especially organized operators, we borrow a lot of these techniques from the financial fraud prevention market, and then ensure that we are applying some of those best practices into the hiring process so that even before a candidate starts an interview, we are able to flag fraudulent actors so you're not exposing yourself to any unforeseen risk, even from an interview standpoint. The second stage, which all of you are, especially if you're a TalView customer, you're already familiar with, is the real-time signal fusion, which we provide. But it's now even better because it uses the living ontology which you have developed as a part of AAOP to fuse all the seven-layer signals into one single incident management dashboard where you get high confidence multi-layer fraud signals and you are not uh being uh bombarded with all the noise. We've actually borrowed some of the best practices from cybersecurity intrusion detection for this, so that you are not having alert fatigue and you're only seeing the signals when it comes to uh reports as a part of the interview security. And the last stage is the post-interview graph analytics and network intelligence, where we are now bringing together all of this data. So if you have uh candidates who have successfully infiltrated your hiring process in the past with fraudulent means, if they have leaked your interview questions, your exam questions, if they have used any third-party services, we're taking some of the best practices from defense intelligence to bring all of that data together and give you a completely uh um, I would say, proactive approach to risk management, where we can help you trace where the risk vectors are within your organization. We can help you dismantle any specific organized hiring fraud which you're being uh being exposed to because somebody is specifically targeting your organization and all of that, with a risk management dashboard, which flags everything which is going on in the hiring process, all the signals which Talgie is collecting on your behalf, whether it's pre-interview, during the interview, and post-interview, and then give you a dashboard which looks at geography-specific lens, function-specific lens, and business unit function uh specific lens, so that you have a view on where you have risk within your hiring process in the organization, what's the level of exposure you have from a risk standpoint, and what proactive actions you can take to mitigate risk. So, for example, you might want to treat a particular hiring process which you're currently treating as low stake, or consider more high stake because you're consistently seeing a lot of fraud attempts in that particular function, or you treat a specific uh group of uh applicants, especially if they're uh group applicants who are using a specific IP or a specific uh device type. You you can group and create specific rules for those set of uh candidates. So they are not able to circumvent any of these security practices which you want to follow from an interview standpoint, and you have complete protection, a complete hiring in uh trust infrastructure, which is protecting you and your organization from all forms of risk, which is associated with uh hiring security and interview security uh threats. And the most important aspect of it, I already spoke about the ontology, and that's the key component here. Uh, the AILP as everything else which Talvi does will is continuously evolving. Uh, there is a significant amount of human override and expert insight to ensure that we are following the best practices when it comes to ethical use of uh AI, uh, the ensuring that there is no bias which is coming into the entire uh incident management mechanism, how the ontology is developed and maintained. Uh, every data point which you're collecting is feeding back into the living ontology. So every time we detect new forms of threats, you will see new forms of incidents within the report. And the detection mechanism itself or entire uh endeavor is to ensure that it's evolving faster than your threat. So the moment we see a particular form of risk with any of our customers or in any part of your organization, the best practices which we are implementing to detect and neutralize that threat is also being made available for every single interview which is happening on the platform. And these models are continuously retrained and adapted to uh specific use cases which you might have. So the entire AIOP ontology encodes the collective purpose of every organization, every human interviewer, and then ensures that all the best practices across all of our uh interview users is being made available for every organization. So there is that cross-organization coordination which we are facilitating, but without using any specific organization data and only relying on metadata to uh make this happen. In 2025, we named the problem of interview security and launched the seven layer security framework to address that. In 2026, with AIOP Assessment Intelligence Operations Platform, we are further strengthening the defense which we have against the evolving threat of interview security. AOP is not just an upgrade, it's a paradigm shift that enables every organization to put their best fight against the evolving threat of candidates using many of this fraudulent mechanism when it comes to the hiring process. So whether it's uh it's a CHR. Or a VP of talent acquisition, or if you are a CIO or a CTO or an IT security leader, or you are a staffing partner or an IT managed services organization, or you are in the TA technology and compliance world, I believe you need to take this threat seriously, which all of you are, that's why you are here attending this summit, but also learn how some of this evolving threat is compromising your organization's security and risking you, and at the same time also understand how an approach like AIOP is going to put you in a better position to compact the risk and protect your organization. The unfortunate reality is that the candidate who probably aced your technical interview that might not be the software engineer who shows up for work on day one. And that's the reality which we need to take into account and prepare to respond to. And it's time to build, whether it's with Talvi or otherwise, to build the trust infrastructure for your organization so that your organization is protected, you are managing risk in the most strong manner. We are going to have an exciting group of sessions today throughout the day. The session one is a look into the IT fraud landscape, what's actually happening in 2026. We have Andrew Stark, who is the CEO of TechCheck Africa. We have Patrick Dunlop, who is a professor at Curtin University, along with Syan Gutta, a lead product manager at Talvue, talking about the IT fraud landscape. In session two, we have some leading practitioners like Ashog Vira at Accenture, Anupam Srivastava from RailTo, and Sumogesty from Dr. Sumogesti from Teke on Co-op. Uh, in a panel moderated by Pratek Shamma from Talview, talking about what organizations are seeing on the ground and what kind of challenges and what kind of approaches they are taking to solving for some of those challenges. And in the last session, we have Praveen Srivastava from ECL Tech, Devenda Kumar from Persistent Systems, and Harish Badia from Conferi in a panel moderated by Sunu Philip from Talview, talking about how leaders are responding, how they're building the answer, building the trust infrastructure on the ground in action, and what are their learnings from their current solutions which they have deployed. So it's going to be a power packed uh summit, a lot of learnings, a lot of insights. Uh, I will see you soon towards the end of the summit, where we regroup to summarize some of the learnings and some takeaways. Thank you.