Thread by @kpoireault

🧵 While #ChatGPT is grabbing the headlines, pushing @GoogleAI (#LaMDA) and @Baidu_Inc to rush their plans to launch competitors, maybe it’s time to explore the security/privacy concerns generative AIs raise.

Here's a #thread on some of these, brought to you by @InfosecurityMag.

@GoogleAI @Baidu_Inc @InfosecurityMag First, @OpenAI’s #ChatGPT has already been used to develop #malware and other malicious tools, as @a_mascellino reported on January 9, 2023. #encryption #cybercrime cc @SShykevich @_CPResearch_
infosecurity-magazine.com/news/chatgpt-d…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ Here’s the original @_CPResearch_ report 👇

https://twitter.com/_CPResearch_/status/1611393636511502338?s=20&t=qVtcTkAVqGwLUlHuBVv6zA

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ Four days later, the same @a_mascellino wrote about another @_CPResearch_ advisory about how #Russian hackers are trying to bypass #ChatGPT’s restrictions for malicious purposes.
infosecurity-magazine.com/news/russian-h…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ Once again, here is the original advisory 👇 blog.checkpoint.com/2023/01/13/rus…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ And it continued, with #ChatGPT allegedly being used to create #polymorphic malware, @CyberArk’s @EranShimony @OmerTsarfati reported on January 17.
infosecurity-magazine.com/news/chatgpt-c…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati Here’s the previously mentioned report 👇
cyberark.com/resources/thre…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati A few days earlier, on January 13, @philmuncaster quoted @PicusSecurity’s co-founder Suleyman Ozarslan (@su13ym4n) and @ExtraHop senior technical manager, @MolesJamie. They both argued that #ChatGPT could, in some aspects, democratize #cybercrime. infosecurity-magazine.com/news/experts-w…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie Our guest columnist Javvad Malik (@J4vv4D), from @KnowBe4, also wrote a post on the @InfosecurityMag’s website where he touched on some #cybersecurity concerns we should have on the use of #ChatGPT. #phishing # socialengineering #scamming cc @_CPResearch_
infosecurity-magazine.com/blogs/chatgpt-…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 But security is not the only area the @InfosecurityMag team investigated on in regards to #ChatGPT…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 On January 24, @ReporterCoker reported on @CiscoSecure’s 2023 Data #Privacy Benchmark Study.

This report showed that 65% of consumers have already lost trust in organizations over their AI practices. 👇
cisco.com/c/en/us/about/…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 @ReporterCoker @CiscoSecure Additionally, @ReporterCoker spoke with @RobertWaitman (@CiscoSecure) @priv_i_see (@bhconsulting) and @GemservCyber on #dataprivacy concerns around the use of tools like #ChatGPT. infosecurity-magazine.com/news/consumers…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 @ReporterCoker @CiscoSecure @RobertWaitman @priv_i_see @bhconsulting @GemservCyber Because it was #DataPrivacyWeek, I also investigated #ChatGPT’s potential privacy issues.

First, @alexanderhanff, @DHillemann (@Fieldfisher) and @kuriharan (@pbdlab) told me how the training model for #ChatGPT was a privacy concern in many ways. #fairuse
infosecurity-magazine.com/news-features/…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 @ReporterCoker @CiscoSecure @RobertWaitman @priv_i_see @bhconsulting @GemservCyber @alexanderhanff @DHillemann @Fieldfisher @kuriharan @pbdlab Then, they also pointed me towards @OpenAI’s lack of transparency in how it was dealing with personal information and data protection regulations. #GDPR #CCPA

infosecurity-magazine.com/news-features/…

@GoogleAI @Baidu_Inc @InfosecurityMag @OpenAI @a_mascellino @SShykevich @_CPResearch_ @CyberArk @EranShimony @OmerTsarfati @philmuncaster @PicusSecurity @su13ym4n @ExtraHop @MolesJamie @J4vv4D @KnowBe4 @ReporterCoker @CiscoSecure @RobertWaitman @priv_i_see @bhconsulting @GemservCyber @alexanderhanff @DHillemann @Fieldfisher @kuriharan @pbdlab By the way, is #ChatGPT compliant with #GDPR? @DHillemann tried to give some elements to answer this question there 👇
fieldfisher.com/en/insights/do…