Exonara Technologies, a Nigeria-based AI and Cybersecurity company, has launched Hacked Before Breakfast, a consumer guide focused on how generative AI is changing cybersecurity risks and the practical measures individuals can take to protect themselves.
The launch responds to a shift already underway. For years, consumer cybersecurity advice has rested on a single, largely unexamined assumption: that a fraudulent call, message, or video carries some detectable flaw — a strange turn of phrase, an unnatural pause, a voice that isn’t quite right. Generative AI has begun to erode that assumption at scale, closing the gap between a genuine communication and a fabricated one to a degree most existing security guidance has not caught up with.
A measurable shift, not a gradual one
The scale of the change is documented. McAfee’s 2025 research recorded a 1,210% year-on-year surge in AI voice cloning scams, and found that 70% of people cannot reliably distinguish a cloned voice from a genuine one — a direct measure of how far instinct-based detection has fallen behind the technology it is meant to guard against. The United States Federal Bureau of Investigation’s Internet Crime Complaint Centre (IC3) reported more than USD 410 million in losses attributed to voice-clone fraud in the first six months of 2025 alone. In South Africa, deepfake fraud rose 269% year-on-year, according to identity verification firm Sumsub.
According to Exonara Technologies, the majority of existing cybersecurity education was built for a threat environment in which a fake could reasonably be expected to look, sound, or read as artificial. Generative AI has narrowed that margin substantially, and Hacked Before Breakfast is structured around defences that do not rely on identifying a fake in real time.
What the guide covers
The guide is organised into ten chapters spanning AI voice cloning, deepfake video calls, phishing, password security, and identity theft, concluding with a structured thirty-day action plan designed to move a reader from awareness to a completed set of protective habits. Its Complete Bundle edition includes a companion workbook containing a printable emergency reference sheet and dedicated space for readers to record a family verification phrase and personal account security checklist.
Localised by market, not translated into it
Each of the guide’s four language editions — English, Deutsch, Español, and Français — was researched independently, built around the institutions its readers would actually engage with in a real incident. The Nigerian edition names the Economic and Financial Crimes Commission (EFCC) and addresses fraud patterns specific to GTBank, Zenith Bank, Opay, and PalmPay. The German edition draws on guidance from the Bundesamt für Sicherheit in der Informationstechnik (BSI) and Verbraucherzentrale. The Spanish edition references INCIBE, and the French edition cites cybermalveillance.gouv.fr, the French government’s official cybersecurity awareness platform.
“Most cybersecurity advice still assumes the person on the other end of an attack can be identified by instinct,” said Precious Oladapo, Founder and CEO of Exonara Technologies. “Generative AI has made that assumption significantly less reliable. The defences that actually hold up now are the ones that do not depend on detecting the fake at all.”
A defence that does not depend on detection
The guide’s central recommendation — a private, pre-agreed family verification phrase used to confirm a caller’s identity during a suspected emergency — is deliberately built around a defence that succeeds without requiring correct, real-time identification of an AI-generated voice or video.
The approach is not exclusive to Hacked Before Breakfast: INCIBE recommends a near-identical method for Spanish consumers, and a separately authored publication, The Scam-Proof Family Playbook, centres its own flagship feature on the same concept under the name “Family Safe Word System.” Exonara Technologies presents this convergence, arrived at independently by separate authors working the same problem, as evidence the technique is effective rather than original to the guide.
The guide additionally offers a technically grounded method for identifying deepfake video calls: asking a caller to turn to a profile angle. Most real-time deepfake generation models are trained predominantly on front-facing footage, a documented limitation of the underlying systems, and rendering quality tends to degrade visibly once that angle changes.
Every statistic cited in the guide carries a named source, listed in full on a dedicated sources page published alongside the guide, including McAfee, the FBI’s Internet Crime Complaint Centre, BSI, Verbraucherzentrale, INCIBE, cybermalveillance.gouv.fr, and Sumsub.
Availability
Hacked Before Breakfast is available now as an Essential Guide or as a Complete Bundle including a companion workbook, across all four language editions. More information is available at hackedbeforebreakfast.com.
About Exonara Technologies
Exonara Technologies is a Nigeria-based AI and Cybersecurity company co-founded by Precious Oladapo and Boluwatife Ajayi, both with backgrounds in AI and cybersecurity. The company is focused on building AI and cybersecurity products.
