In recent years, the telephone has become a gateway to countless scams, from impossible offers to fake technicians looking for your data. In this context, an idea as ingenious as it is effective has burst onto the scene: a virtual grandmother named Daisy who, with infinite patience and a lot of talk, entertains scammers for long minutes so that they don't bother real users and to expose their tactics. This tool, promoted by Virgin Media O2, demonstrates that technology can also work in favor of the consumer when it is designed with purpose and a touch of mischief, and it does so based on Credible conversations, quick responses, and no fear of boring the scammer..
The picture of the problem is clear: according to data shared by O2, a significant portion of the population frequently receives fraud attempts and there is widespread concern about the possibility of falling for one. Faced with this scenario, Abuela Daisy appears as an active shield that wastes criminals' time and, in the process, documents their methods to improve prevention. We are talking about an AI that sounds like a loving grandmother, that can talk about sewing, family, or grandchildren, and that even provides fabricated information to simulate a real victim, so that the scammer falls into their own trap and has no room to annoy other people, a strategy of automated scambaiting at scale.
What is Grandma Daisy and why does she exist?
Daisy is a conversational artificial intelligence created by Virgin Media O2 to interact with phone scammers and keep them busy as long as possible. The company introduces her as the newest member of its anti-fraud team, with a role as descriptive as it is amusing: Director of Scam Relations, reflecting her core mission.
The goal is not only to entertain criminals to the point of boredom, but also to record their patterns and point out to users what signs reveal a scam, thus enabling more people to protect themselves and fewer victims to be exposed to these scams, something that the brand includes in its Swerve the Scammers initiative, focused on awareness, prevention and response.
The project's genesis also responds to a social impulse. Internal research indicates that a vast majority of citizens would like to get on the scammers' level and get back at them, but don't have the time to do so. Enter Grandma Daisy with her boundless patience and affectionate nature, which serves to prolong conversations to unimaginable limits and make the scammers believe they're easy prey.
Her friendly grandmotherly tone is no coincidence, as it exploits the bias with which criminals often target older people. In practice, this AI transforms that idea into a defensive weapon, turning a supposed victim into an inexhaustible adversary who, moreover, never gives out real data or makes security mistakes.
This is how O2's AI grandmother works
Under the hood, Grandma Daisy combines multiple generative models to string together realistic responses and maintain the thread of the conversation in real time. O2 and AI agency Faith, part of the VCCP group, have customized a large language model—a process that requires learn to program from scratch—to give her personality, memory, and a repertoire of stories that seem real. Her voice has been synthesized based on that of a real grandmother on the team, seeking naturalness and credibility, while her appearance for the advertising campaign has been created with a specifically trained broadcast model. The whole result is a photorealistic audiovisual presence and a fluid conversation that, to the scammer's ears, is indistinguishable from a person, with the advantage that the AI is capable of answer, improvise and redirect the dialogue in a matter of seconds.
To fine-tune Daisy's behavior, creator Jim Browning, a leading figure in the scambaiting community for his work exposing fraudulent call centers online, was involved. Based on his experience, the AI learned not only to speak credibly but also to use subtle distractions, pauses, personal stories, and fictitious data to sustain the deception without putting anyone at risk.
The result is a system that can handle calls of more than 40 minutes, in which the grandmother talks about everything, is interested in domestic issues and gives the runaround every time the scammer tries to close the bait, something that in operational terms means that the malicious person stops bothering real users, while at the same time a useful knowledge base is being fed for Reinforce filters and improve preventive blocking.
O2 has been testing Grandma Daisy in real-world settings for weeks, in conversations without human intervention, with striking results. In its early stages, it handled more than a thousand interactions since its launch, with many conversations lasting tens of minutes and explicit expressions of frustration from scammers.
This deployment allows the teleoperator to measure what types of hooks the criminals are using, what doubts are most frequent and at what times they insist on asking for banking data, codes or remote access, vital information to improve spam detection systems and filter suspicious contacts, in addition to serving as educational material so that customers know how to identify warning signs and hang up in time without falling into deception.
Most common phone scams you're trying to stop
vishing
Vishing is phone phishing. The scammer poses as a bank employee, energy operator, or telecommunications technician and requests sensitive information under a reasonable-sounding pretext, such as verifying a charge or preventing a service outage. The real goal is to extract information to attack an account or authorize a payment.
For any call of this type, remember that entities will not ask you for passwords or one-time codes over the phone and that you should verify the contact through official channels before saying anything, because The bank will never confirm passwords or full numbers on an incoming call..
Double call
A classic that still works. First, someone claiming to be your utility company contacts you to inform you of an urgent rate increase or a billing problem. Minutes later, a supposed competitor calls you to rescue you with a miraculous offer. They take advantage of the comparison effect to get you to accept the change.
No reputable company will notify you of price increases through this means or force you to sign a contract. If this happens to you, hang up, find your supplier's official number, and confirm this yourself. This sudden pressure is an indicator of fraud, and the argument for the increase is usually a scam. hook designed to force an impulsive decision.
Phantom calls
You pick up and no one speaks. This is common with automated systems that validate which numbers are active or with campaigns that try to get you to return the call. Don't say anything, hang up immediately, and block the contact. There are abusive practices that, when you answer, may attempt to divert you to premium-rate services or use your response to target you for future actions.
Do not feed the scammer's blacklist and avoid any interaction, since in this type of maneuvers your voice or your simple availability can be reused against you, one more reason to mark the number as spam and move on.
Voice recording
Another common tactic is to provoke you into saying yes to an innocent question, with phrases like "are you listening" or "do you confirm?" This "yes" can be truncated and reused to authorize supposed contracts or positions. Avoid responding affirmatively to unsolicited calls and always demand information in writing. If you're suspicious, hang up and verify. And remember that edited recordings can fit your voice into false contexts, so caution is a powerful ally against this form of scam based on social engineering and audio manipulation.
Eavesdropping
Fake technicians who ask to install remote assistance software are a major danger. They claim they need to check your computer or phone, and after convincing you, they gain access to your devices, your calls, and your data. No reputable company will ask you to install tools outside of their official channels or pressure you to give up control.
If a stranger persists, end the call and consult on your own. Installing dubious applications opens the door to spying, password theft, and information hijacking, so the rule is clear: never install anything a stranger asks you to on the phone, especially if they try to take advantage of your urgency or ignorance.
Answering machines
There are campaigns that, using bots, request your information to manage prizes, refunds, or supposedly limited offers. They'll ask for your full name, address, card number, or IBAN. The real intention is to gather information with commercial or criminal value. In practice, no reliable promotion is managed without verification and transparent information.
Reject these requests, don't press any keys if asked, and hang up. Also, add the number to your blocked list and, if you're a customer of an operator with identification services, activate them so you're notified when there's a risk of malicious robocall.
Lost calls
Also known as wangiri calls, these consist of a single or double tap from a strange, usually international, prefix to get you to call back out of curiosity. In doing so, they connect you to premium-rate numbers or keep you on hold, incurring charges. Prevention is simple: don't return calls from unknown numbers with strange prefixes, and use apps or carrier services to identify the source.
If in doubt, look up the prefix and check user forums for spam reports, as the goal of this scam is to get you to inadvertently pay for a call you didn't request, a simple trick that still works if you don't exercise extreme caution.
Psychology and tactics of scammers
Criminals play with time, pressure, and emotion. They want you to make quick decisions, to avoid thinking, and to feel a sense of fear of losing something or a sense of urgency to avoid a problem. Therefore, they demand immediate action, discretion, and promise to resolve everything if you give them one more piece of information. Grandma Daisy neutralizes this dynamic with friendly conversation, patience, and rambling, which defuses the urgency, disrupts the scammer's rhythm, and forces them to dedicate resources to no benefit.
This investment of time reduces opportunities to contact other potential victims and also helps identify what scripts and arguments they use, a valuable input for automatic filters and blocking systems that take care of the client in the background. more context and better signals.
The frustration many scammers display when they encounter Daisy demonstrates that their business model depends on maximizing the number of calls and closing impulsive transactions. If each conversation lasts half an hour, business suffers. Hence, this AI serves a dual purpose: active defense and generating anti-fraud intelligence.
As an added effect, this material becomes practical education for users, who learn to distrust rushing, to recognize recurring voices or scripts and not to give in when someone promises the moon and the stars in exchange for verification, because a hook presented as an opportunity usually hides the request for key information, and that is the point at which you have to say enough is enough and hang without remorse.
What is O2 doing besides Grandma Daisy?
The company has strengthened its protection ecosystem with artificial intelligence systems that block spam and detect anomalous patterns in calls and messages. It also offers free caller ID services for customers, so they can more clearly see who's calling and whether there's a potential risk. It has also enabled a very simple short number, 7726, to which users can forward suspicious calls or text messages.
These referrals help train the models and cut off active campaigns, creating a virtuous circle between users and operators, with quantifiable results in a short time, such as the blocking of tens of millions of messages and the interception of suspicious operations for hundreds of millions of pounds that, otherwise, could have ended up in accounts of organized gangs.
This technical approach is complemented by a clear public call to the UK authorities to step up the fight against fraud. O2 advocates for the appointment of a specific government figure dedicated to this issue and the creation of a single, specialized, and well-resourced law enforcement agency to centralize the investigation of cases.
Fragmentation and lack of coordination favor criminals, so a single structure would facilitate cooperation, streamline responses, and make it difficult for fraudsters to exploit gaps between jurisdictions. The idea is that protection does not depend solely on the end user or a specific operator, but rather that there is a robust national framework that makes this type of crime more costly and risky, supported by clear rules and effective enforcement of the law.
Swerve the Scammers Campaign Results
To raise awareness, O2 has launched a creative campaign starring Daisy and supported by the VCCP agency through its artificial intelligence division, Faith. The audiovisual piece and associated communications show the AI grandmother's real-life interactions with scammers, including lengthy conversations about family life, sewing, and everyday life, and moments in which the AI enters fabricated data to keep the scam under control.
The intention is both educational and dissuasive, since seeing the scammers lose patience and leave empty-handed serves to demonstrate that there are ways to take away their time and room for maneuver.
The campaign also featured well-known faces, such as Amy Hart, who shared her personal experience after suffering bank fraud. Her testimony highlights the real impact of these scams and reinforces the need to be more cautious when using the phone, while reminding everyone that no one is immune to falling for the scam if the conditions of the scam are well planned.
Another key to communication has been giving voice to a recurring social desire: the idea of getting back at scammers without having to dedicate hours to it. Here, Grandma Daisy becomes the ideal ally, assuming that role with tireless professionalism and giving customers confidence that, if they report and block, there are technical resources working for their safety. background and large scale.
How you can protect yourself on a daily basis
Beyond Grandma Daisy, there are habits that make a difference. If a call asks for sensitive information, hang up and contact them yourself through official channels. Don't share passwords, one-time codes, full card numbers, IBANs, or login details.
Be wary of emergencies, unrefusable offers, and unexpected prizes. And remember, if a company needs your confirmation for something serious, they will offer verifiable alternatives, such as going to the official app or an office. The first filter is you, and that filter works best with a cool head, common sense, and the ability to say no in time, a gesture that's enough to deactivate the vast majority of attempts. social engineering.
- Activate your carrier's caller ID and, if desired, complement it with well-known number reputation solutions, similar to Truecaller, to help you recognize prefixes and suspicious sources, especially when unusual international calls appear or with repetitive patterns that indicate spam, an additional layer that adds to the protection of the own operator.
- Report suspicious calls and text messages to the operator's channels in your country. In the United Kingdom, O2 has popularized the number 7726 for sending spam. This collaboration improves detection systems and prevents others from falling prey. In Spain, you should also sign up for advertising exclusion lists like the Robinson List to minimize unwanted legitimate marketing, although its effectiveness against pure fraud is limited, because criminals don't respect the rules, and that's where it's convenient. block and report.
- Don't install software at the request of strangers or grant remote access to anyone who calls you unsolicited. If you suspect malware or a technical failure, open your bank or provider's official app, or go to certified and verified support, avoiding links received through unofficial channels and always keeping your system updated, a basic step for close doors to intruders.
- Avoid answering "yes" or providing personal information to unsolicited calls. If you're asked to confirm your identity, request a secure alternative channel or go to the official app or website. Don't return calls to unknown numbers with strange prefixes, and use an internet search to see if other users have reported that phone as dangerous. This simple gesture saves trouble and helps cut down on calls. chains of fraud.
It's also key to review the privacy permissions you grant when registering for services, stores, and sweepstakes. Often, without realizing it, we allow commercial communications by accepting extensive terms and conditions that we don't read. This consent can lead to legitimate calls, which aren't criminal, but they can be overwhelming and open the door to confusion.
Take a few minutes to adjust your preferences and avoid sharing your number unless absolutely necessary. This will reduce the noise and expose clearly malicious attempts, making it easier to detect the scam and stop it with a simple gesture. hang up and lock.
Broader overview: regulation, trends, and recent cases
The regulatory environment is shifting to contain fraud and strengthen cybersecurity. In the United Kingdom, measures have been announced to raise the bar on security for connected devices, including stricter requirements such as banning weak passwords from the factory in the Internet of Things ecosystem.
While this may make certain devices more expensive, the basic security improvement reduces the attack surface and benefits all users. These types of advances, combined with the efforts of operators and manufacturers, mean that fraud must reinvent itself, and every small barrier leads to a real decrease in opportunities for criminals. This line of work complements practical initiatives like Daisy and the tools for smart lock.
The sophistication of deception also comes with artificial intelligence technologies used by criminals themselves, such as voice cloning and deepfakes. A notable case was the attempted scam that affected the CEO of a large advertising group, demonstrating that even well-resourced companies can find themselves in trouble if the attack is well planned.
This context reinforces a key message: trust verification through official channels, be wary of incoming calls demanding immediate action, and establish internal procedures in businesses and homes to validate payments, account changes, or transfers. Education, block lists, and call analysis tools form a highly effective triangle to keep the threat at bay and reduce the effectiveness of these attempts. advanced impersonation.
Finally, it's worth remembering that information and ongoing attention make the difference when it comes to avoiding scams. Media outlets and platforms are promoting channels to bring news and security tips to users, and many newsrooms have created alert spaces in their apps and messaging channels to disseminate best practices.
Staying up to date and sharing tips with family and friends, especially older adults, multiplies the preventative effect, and when an operator offers free identification tools or specific reporting numbers, taking advantage of them is equivalent to doing your part in a community defense that not only protects you, but also adds friction to scammers and strengthens the ecosystem from the ground up. collective security.
Daisy embodies a powerful idea: putting artificial intelligence on the consumer's side with creativity and rigor. By boring scammers, she slows down their campaigns and leaves them breathless, while teaching citizens to recognize the typical signs of fraud. Combined with automatic blocking, caller ID, and active user collaboration through reporting and privacy settings, Grandma AI fits into a broader strategy that combines technology, education, and regulation.
And while no system is foolproof, every minute a scammer wastes with Grandma Daisy is one less chance someone will fall for it, a small but valuable victory that, multiplied by thousands of calls, becomes a real wall in the face of nuisance calls and telephone scams. Share this guide and help others learn about Grandma Daisy and how to use it..