Key takeaways
- Binance hack vulnerabilities: Reverse engineering and social engineering were central to the US$6.3bn Binance cryptocurrency hack in October 2026.
- Social engineering in Binance hack: Hackers used a fake job interview to encourage a Binance software engineer to download malicious software, hidden in an interview assignment.
- Cyber vulnerabilities in recruitment: Recruitment platforms and processes are often targets for cyber criminals, from malware hidden in employment documents and applications to hacks of recruitment platforms, such as the Healthdaq attack by hacking group XP95.
- Social engineering awareness: Companies must adopt exercises that simulate modern social engineering techniques rather than traditional phishing emails, advises Dr Emma Arakas.
In the cyber attack on cryptocurrency exchange Binance, hackers made off with some $6.3bn (£4.7bn) worth of its holdings in under 30 minutes – the largest heist in history.
Although the actual theft took less than a half-hour to complete, the thieves took their time initiating the overall plan, which included the use of a simple fake job placement.
As far back as June, a Binance system engineer downloaded some malicious AI-driven software. To the company’s credit, it acted quickly and wiped the infected computer clean.
But the software had burrowed itself deep into sensitive areas of the Binance enterprise system. It began learning about the target’s user profile and habits, in a round of ‘reverse engineering’ fast becoming popular among hackers in learning all they need to know about a system’s inner workings, including the routines and vulnerabilities of its users.
This highlights the risks that recruitment can pose for organisations, both from potential candidates and employees looking for new opportunities.
How hackers broke into Binance
This is how the hackers got acquainted with their mark, well enough and long enough to invite him to a fake job interview.
“Software Engineer B,” according to Binance, was contacted through “a legitimate job-hunting platform” by someone posing as a project director, who had seen the engineer’s profile referencing his experience at Binance. Under the guise of an interview assignment, the engineer was convinced to download malicious software.
The crypto robbery had all the makings of an old-fashioned bank heist, where the thieves would recce the target in advance, track staff when they come into work and when they leave, says Danilo Poccia, Director of AI Solutions at Ireland-based agentic AI consultancy Nearform. It is very much the same principle, updated for the age of AI.
“Essentially, enterprises put a lot of effort into securing the technology, and less into securing how people, processes and technology interact, and now how AI agents interact with all three. That is where this attack got through,” he explains.
Common recruitment-related attacks
Recruitment is a vulnerable pathway into many organisations, usually targeted via well-hidden phishing attacks. Attackers will hide as genuine-seeming recruiters or candidates, using malware hidden in documents and applications typically exchanged in recruitment processes, such as CVs.
Attackers also target recruitment platforms and software as part of ransomware attacks. In March 2026, healthcare recruitment platform Healthdaq was attacked by hacking group XP95, which claimed to have stolen half a million files relating to staff and candidates.
Job advertisements can also give away pertinent details about an organisation, such as leadership changes, heads of department, and investments the organisation is making.
As such, it’s important for organisations to ensure their recruitment processes are as secure as possible, and that individual staff know the risks associated with recruitment and job sites.
Creating safeguards: limit systems access
The common refrain from cyber security experts is that employee access to enterprise systems must be limited.
“Employees should never be given so much privileged access that it can snowball to become the biggest heist in history,” says Jake Moore, Global Cybersecurity Advisor at ESET. “Companies need to restrict employee access to sensitive systems or abilities to move huge amounts of cash and crypto without sounding alarm bells.”
“Even legitimate employee accounts can be compromised, so detecting suspicious activity early is vital. But unfortunately, newer banking organisations and crypto companies often favour speed over security,” he adds.
Remove long-term access credentials
Poccia at Nearform links the problem directly to “long-term access credentials”, an enduring enterprise technology feature that now looks more like a gaping liability in 2026, and which in all probability gave the hackers the five months they had to create and finesse their plan.
“Always use short-term credentials, especially on devices,” is Poccia’s advice. “This way, if a credential is compromised, you’ve only minutes or hours before they expire. It’s easier to react quicker and hackers have less time to plan something as sophisticated [as the Binance heist]. And of course, use multi-factor authentication or a second confirmation, on a separate device, for any action that moves money.”
Maximise awareness training
As the recent National Cyber Security Centre’s 2026 Annual Cyber review points out, social engineering is still a huge issue for firms, so organisations need to do more to address it.
For psychologist Dr Emma Arakas, Principal Consultant, Applied Human Sciences at Frazer-Nash Consultancy, the practical lesson from the heist “is that organisations need to move beyond awareness training that merely focuses on spotting suspicious emails”.
Instead, she says, organisations should:
- train staff to recognise social engineering scenarios, including recruitment approaches, collaboration requests and networking opportunities;
- establish clear rules for handling interview exercises, coding tests and external repositories on corporate devices;
- encourage a “psychologically safe culture”, where employees can pause and seek advice without fear of appearing obstructive or incompetent;
- limit the impact of a compromised account or device through strong segregation of privileges and robust endpoint monitoring; and
- use “realistic exercises” that simulate modern social engineering techniques rather than traditional phishing emails.
“Most importantly, organisations should assume that determined attackers will eventually establish contact with employees,” Arakas advises.
“The Binance case is a reminder that sophisticated cyber attacks increasingly resemble influence campaigns as much as technical intrusions. For this reason, understanding how trust is built, exploited and defended is becoming just as important as understanding the technology itself.”
Cyber security awareness
View a series of resources and a podcast addressing the latest cyber security issues and how to protect your business.
- What Binance hack tells us about cyber recruitment vulnerabilities
- How financial criminals – and their pursuers – are turning to blockchain
- AI and client data: complying with the principle of confidentiality
- Cyber attacks in retail: from prevention to resilience
- How to use AI to strengthen your cyber defences
Stay up to date
You can receive regular email updates from ICAEW insights, including weekly or monthly enewsletters. Subscribe to whichever works for you.
Sign up