Episode 291 of the Transatlantic Cable Podcast kicks off with news around OpenAI’s recent decision to reduce the amount of information it’s sharing about how ChatGPT is trained, causing some to worry that it’s no longer as open as originally designed. 

From there discussion moves to a recent story around the DEA (Drug Enforcement Administration) in America and hackers looking to leverage DEA’s internal toolset.  Following that, news around hacked crypto ATMs and a new story from Securelist around  a newly discovered APT group, CommonMagic. 

This episode also includes a discussion with Susi O’Neil from Kaspersky’s Brand Activation Studio to talk about their upcoming audio mini-series called “Insight Story.” The series aims to help business leaders better understand digital tech such as AI, Metaverse and much more.  To listen to episode one, you can tune in here. 

If you liked what you heard, please consider subscribing. 

·        OpenAI co-founder on company’s past approach to openly sharing research: ‘We were wrong’

·        Inside the DEA Tool Hackers Allegedly Used to Extort Targets

·        General Bytes Bitcoin ATMs hacked using zero-day, $1.5M stolen

·        Bad magic: new APT found in the area of Russo-Ukrainian conflict

Please accept our apologies for the audio in places - we've done what we can to clean it up - back to normal scheduling next week! 

Welcome to the 290^th episode of the Kaspersky Transatlantic Cable podcast. In this episode, we kick things off with a discussion on how ransomware gangs are taking new leaps to extort money. Spoiler alert - you aren’t going to like it one bit. From there, we jump into a discussion surrounding the booming black market in China for access to ChatGPT. 

We continue the discussion looking at how a fitness app led a researcher to discover the home and walkabouts of a former top aid to Joe Biden. To close out the episode, we look at how banks’ voice as a password is not secure. 

If you liked what you heard, please consider subscribing. 

·        Ransomware Attacks Have Entered a ‘Heinous’ New Phase

·        A booming illicit market for OpenAI’s chatbot shows the huge potential, and risks, for Chinese generative AI

·        AllTrails Data Exposes Precise Movements of Former Top Biden Official

·        How I Broke Into a Bank Account With an AI-Generated Voice

In episode 289 of the Transatlantic Cable, the team look at four new stories to tempt your earbuds.  This week kicks off with news that TikTok are implementing a 60-minute time limit for users under 18. 

Following that, discussion moves to further developments around FTX, this time about their missing $9 billion. To wrap up, we discuss news around Call of Duty players taking back Black Ops III from hackers and bots and more news around the recent LastPass data breach.

If you liked what you heard, please consider subscribing.

• TikTok sets 60-minute daily screen time limit for under-18s
• FTX Confirms $9 Billion in Customer Funds Vanished
• Gamers are fixing a video game ‘taken over’ by hackers
• LastPass says employee’s home computer was hacked and corporate vault taken

The latest episode of the Transatlantic Cable sees the team starting out with news around Signal and their refusal to weaken encryption for the U.K government.  Following that, the team move onto discussions around Meta and National Center for Missing and Exploited Children’s (NCMEC) campaign to help people with the spread of minors' intimate images online.

To wrap up, the team discuss SnapChat A.I chat bots and LockBit’s attack against the U.K’s Royal Mail service.

If you liked what you heard, please do consider subscribing. 

·        Signal would 'walk' from UK if Online Safety Bill undermined encryption

·        LockBit leaks 44GB of Royal Mail's data and sets fresh £33 million ransom

·        Meta supports new platform preventing spread of minors' intimate images online

·        Snapchat is adding OpenAI chatbot capabilities for the new My AI feature

The latest episode of the Transatlantic Cable starts off with news that GoDaddy, one of the largest web hosting services was hacked back in 2020 and the attackers were able to gain access to the back end of millions of websites, helping them to push malware and steal credentials.

Moving from there the team look at news of the recent remove of SMS 2FA (two factor authentication) from Twitter, sparking a fierce debate in the infosec world. There’s also discussion around spam bots in Instagram and the (final) death of Internet Explorer 11.

The team was also lucky enough to chat to David Emm, principle security researcher at Kaspersky about new research around kids attitudes to fraud and security online. 

If you liked what you heard, please do consider subscribing.

• GoDaddy says a multi-year breach hijacked customer websites and accounts
• Twitter to charge users for text-message authentication
  
• Why Are ‘Porn Stars’ Liking Your Instagram Stories?
  
• RIP Internet Explorer: Microsoft Kills Off Legacy Browser

The latest episode of the Transatlantic Cable kicks off once again with a discussion around A.I – this time, looking at how it could soon be used by musicians to help them create music – but concerns grow that it could one day be used to create music (much like it can now be used to create art). From there the team discus how a pair of developers have created an A.I that generates lifelike police portraits sketches, and it’s got privacy and civil liberty experts worried. 

Moving away from A.I to something more concrete, the team talk about the recent proposal from the U.K government to ban the use and sale of encrypted phones – often used by criminals to circumnavigate prying eyes.

• David Guetta says the future of music is in AI
• Developers Created AI to Generate Police Sketches
• UK Proposes Making the Sale and Possession of Encrypted Phones Illegal
• Reddit Confirms It Was Hacked—Recommends Users Set Up 2FA

Episode 285 of the Transatlantic Cable kicks off with yet more news around A.I and the ever-interesting topic of ChatGPT.  However, this time, the discussion starts with regulation and the CTO of Open AI believes that it’s inevitable.  Moving from that to a more quirky story, still involving A.I however: a small team have used AI to create a never-ending Seinfield episode, streaming over on Twitch – sadly at the time of writing, it appears to have been shut down, but it still raises interesting questions around content creation, human creativity and what A.I will bring (or take away) from the table. 

Moving away from A.I to more cyber-security related news – the team discuss the closure of Exclu, an encrypted mobile phone company that enabled criminals to chat without fear of snooping, however the Dutch police were already inside the systems.

To wrap up, the team talk about a recent story around the FBI’s changing approach to ransomware, with the recent Hive takedown.

If you liked what you heard, please do consider subscribing. 

·        The Creator of ChatGPT Thinks AI Should Be Regulated

·        Dutch Police Read Messages of Encrypted Messenger 'Exclu'

·        AI has been generating an endless Seinfeld episode

·        Hive takedown illustrates FBI’s evolution towards victim-recovery efforts

Episode 284 kicks off with a story that is going to surprise no one: over 50% of smart appliances are not connected to the internet, which is a surprise to only the manufacturers – it seems people aren’t really convinced by Wi-Fi enabled refrigerators or washing machines, it seems.

Moving on from that, the team discuss news that AI audio is becoming a problem, with AI company ElevenLabs saying that they’re going to start clamping down on troublesome audio recordings – but has Pandora’s box already been opened?  Staying with AI, the team also look at a story breaking from China, with Baidu set to launch their own version of ChatGPT.

To wrap up, the team discuss a rather odd story involving pet fish, Pokemon, PayPal and credit cards…. We’ll leave the rest up to your imagination.

If you liked what you heard, please do consider subscribing.

• Appliance makers sad that 50% of customers won’t connect smart appliances
• AI Voice Firm Clamps Down After 4chan Makes Celebrity Voices for Abuse
• China's biggest search engine is to set launch a ChatGPT rival in March

Pet fish playing Nintendo Switch run up bill on owner’s credit card

Episode 283 kicks off with the recent (good news) story that ransomware gangs earnings are down by quite a bit from last year – turns out that people are starting to treat backups properly and are refusing to pay – great news for 2023!  Following that, the team look at troubling news, this time from the crypto-currency world, with a story looking at the recent collapse of Genesis.

Moving from crypto to the world of fake news, a recent study has shown that the reason people share fake news is perhaps more complex than previously thought. To wrap up, the team discuss concerning news around ChatGPT and exams, with a professor running the AI through an exam for Medical Licenses and Business Degrees and it turns out it can pass them both.  Should teachers and professors be worried? Time will tell it seems.

If you liked what you heard, please do consider subscribing.

1. Cyber-crime gangs' earnings slide as victims refuse to pay
2. Crypto lender Genesis files for bankruptcy
3. Study reveals the key reason why fake news spreads on social media
4. ChatGPT Is Passing Tests Required for Medical Licenses & Business Degrees

Episode 282 kicks off with new research from Which? Showing that ‘smart’ technology might not be as smart as we’ve been led to believe – their research shows that most devices stop receiving updates from 2 years, even thought the device has a suggested life of 10 – it seems more work needs to be done.  From there the team look at two stories around A.I, the first on the media sensation, ChatGPT and how some people are looking to use it for malware creation, and the second looking at a chat A.I program that has become (and I’ll quote from the article here) “too horny”.

In the middle of all this, I sat down with Evgeny Goncharov, head of ICS CERT at Kaspersky to talk to him about upcoming ICS APT predictions for 2023.

Following that, the team discuss news that Norton LifeLocker password manager has been breached, those affected are being urged to update and change their passwords as quickly as possible. 

If you liked what you heard, please do consider subscribing.  

• ICS cyberthreats in 2023 – what to expect
• ICS cyberthreats in 2023 – what to expect Securelist report
• Smart appliances could stop working after two years, says Which?
• Armed With ChatGPT, Cybercriminals Build Malware And Plot Fake Girl Bots
• Replika Users Say the Chatbot Has Gotten Way Too Horny
• NortonLifeLock warns that hackers breached Password Manager accounts

The first episode of 2023 of the Transatlantic Cable Podcast kicks off with news around the “recent” Twitter hack – emphasis on recent as the data was actually siphoned from Twitter back in 2021, but has only recently come to light – timing is everything it seems. 

From there, the team look at news that the Web3 industry lost a whopping $4 billion to crypto-fraud last year – apparently, this is good news as the previous year it was a colossal $8 billion. Staying with crypto, the next story dives into the recent news around the closure of CryptoZoo – a game created by internet star, Logan Paul. The premise of the game allowed users to ‘breed and share’ unique NFT creatures, all whilst earning cryptocurrency.

To wrap up, the last story looks at more news around crypto, this time looking at news that U.S. bank Silvergate saw $8 billion in crypto withdrawals, most near the tail end of 2022. 

If you liked what you heard, please do consider subscribing.

• More than 200 million Twitter users’ email addresses leaked by hackers
• The Web3 industry lost billions to crypto fraud last year
• YouTube star Logan Paul apologizes for CryptoZoo project failure
• US bank Silvergate hit with $8bn in crypto withdrawals