#FactCheck - Debunking the AI-Generated Image of an Alleged Israeli Army Dog Attack

Introduction 

AI has revolutionized the way we look at growing technologies. AI is capable of performing complex tasks in fasten time. However, AI’s potential misuse led to increasing cyber crimes. As there is a rapid expansion of generative AI tools, it has also led to growing cyber scams such as Deepfake, voice cloning, cyberattacks targeting Critical Infrastructure and other organizations, and threats to data protection and privacy. AI is empowered by giving the realistic output of AI-created videos, images, and voices, which cyber attackers misuse to commit cyber crimes. 

‍

It is notable that the rapid advancement of technologies such as generative AI(Artificial Intelligence), deepfake, machine learning, etc. Such technologies offer convenience in performing several tasks and are capable of assisting individuals and business entities. On the other hand, since these technologies are easily accessible, cyber-criminals leverage AI tools and technologies for malicious activities or for committing various cyber frauds. By such misuse of advanced technologies such as AI, deepfake, and voice clones. Such new cyber threats have emerged. 

‍

What is Deepfake?

Deepfake is an AI-based technology. Deepfake is capable of creating realistic images or videos which in actuality are created by machine algorithms. Deepfake technology, since easily accessible, is misused by fraudsters to commit various cyber crimes or deceive and scam people through fake images or videos that look realistic. By using the Deepfake technology, cybercriminals manipulate audio and video content which looks very realistic but, in actuality, is fake. Voice cloning is also a part of deepfake. To create a voice clone of anyone's, audio can be deepfaked too, which closely resembles a real one but, in actuality, is a fake voice created through deepfake technology. 

‍

How Deepfake Can Harm Organizations or Enterprises?

• Reputation: Deepfakes have a negative impact on the reputation of the organization. It’s a reputation which is at stake. Fake representations or interactions between an employee and user, for example,  misrepresenting  CEO online, could damage an enterprise’s credibility, resulting in user and other financial losses. To be protective against such incidents of deepfake, organisations must thoroughly monitor online mentions and keep tabs on what is being said or posted about the brand. Deepfake-created content can also be misused to Impersonate leaders, financial officers and officials of the organisation.
• Misinformation: Deepfake can be used to spread misrepresentation or misinformation about the organisation by utilising the deepfake technology in the wrong way. 
• Deepfake Fraud calls misrepresenting the organisation: There have been incidents where bad actors pretend to be from legitimate organisations and seek personal information. Such as helpline fraudsters, fake representatives from hotel booking departments, fake loan providers, etc., where bad actors use voice clones or deepfake-oriented fake video calls in order to propose themselves as belonging to legitimate organisations and, in actuality, they are deceiving people.

‍

How can organizations combat AI-driven cybercrimes such as deepfake? 

• Cybersecurity strategy: Organisations need to keep in place a wide range of cybersecurity strategies or use advanced tools to combat the evolving disinformation and misrepresentation caused by deepfake technology.  Cybersecurity tools can be utilised to detect deepfakes. 
• Social media monitoring: Social media monitoring can be done to detect any unusual activity. Organisations can select or use relevant tools and implement technologies to detect deepfakes and demonstrate media provenance. Real-time verification capabilities and procedures can be used. Reverse image searches, like TinEye, Google Image Search, and Bing Visual Search, can be extremely useful if the media is a composition of images.
• Employee Training: Employee education on cybersecurity will also play a significant role in strengthening the overall cybersecurity posture of the organisation.

Conclusion

There have been incidents where AI-driven tools or technology have been misused by cybercriminals or bad actors.  Synthetic videos developed by AI are used by bad actors. Generative AI has gained significant popularity for many capabilities that produce synthetic media. There are concerns about synthetic media, such as its misuse of disinformation operations designed to influence the public and spread false information. In particular, synthetic media threats that organisations most often face include undermining the brand, financial gain, threat to the security or integrity of the organisation itself and Impersonation of the brand’s leaders for financial gain.

Synthetic media attempts to target organisations intending to defraud the organisation for financial gain. Example includes fake personal profiles on social networking sites and deceptive deepfake calls, etc. The organisation needs to have the proper cyber security strategy in place to combat such evolving threats. Monitoring and detection should be performed by the organisations and employee training on empowering on cyber security will also play a crucial role to effectively deal with evolving threats posed by the misuse of AI-driven technologies. 

References:

• https://media.defense.gov/2023/Sep/12/2003298925/-1/-1/0/CSI-DEEPFAKE-THREATS.PDF
• https://www.securitymagazine.com/articles/98419-how-to-mitigate-the-threat-of-deepfakes-to-enterprise-organizations

‍

Introduction

The use of digital information and communication technologies for healthcare access has been on the rise in recent times. Mental health care is increasingly being provided through online platforms by remote practitioners, and even by AI-powered chatbots, which use natural language processing (NLP) and machine learning (ML) processes to simulate conversations between the platform and a user. Thus, AI chatbots can provide mental health support from the comfort of the home, at any time of the day, via a mobile phone.  While this has great potential to enhance the mental health care ecosystem, such chatbots can present technical and ethical challenges as well.

Background

According to the WHO’s World Mental Health Report of 2022, every 1 in 8 people globally is estimated to be suffering from some form of mental health disorder. The need for mental health services worldwide is high but the supply of a care ecosystem is inadequate both in terms of availability and quality. In India, it is estimated that there are only 0.75 psychiatrists per 100,000 patients and only 30% of the mental health patients get help. Considering the slow thawing of social stigma regarding mental health, especially among younger demographics and support services being confined to urban Indian centres, the demand for a telehealth market is only projected to grow. This paves the way for, among other tools, AI-powered chatbots to fill the gap in providing quick, relatively inexpensive, and easy access to mental health counseling services.

Challenges

Users who seek mental health support are already vulnerable, and AI-induced oversight can exacerbate distress due to some of the following reasons: 

1. Inaccuracy: Apart from AI’s tendency to hallucinate data, chatbots may simply provide incorrect or harmful advice since they may be trained on data that is not representative of the specific physiological and psychological propensities of various demographics. 
2. Non-Contextual Learning: The efficacy of mental health counseling often relies on rapport-building between the service provider and client, relying on circumstantial and contextual factors. Machine learning models may struggle with understanding interpersonal or social cues, making their responses over-generalised. 
3. Reinforcement of Unhelpful Behaviors: In some cases, AI chatbots, if poorly designed, have the potential to reinforce unhealthy thought patterns. This is especially true for complex conditions such as OCD, treatment for which requires highly specific therapeutic interventions. 
4. False Reassurance: Relying solely on chatbots for counseling may create a partial sense of safety, thereby discouraging users from approaching professional mental health support services. This could reinforce unhelpful behaviours and exacerbate the condition. 
5. Sensitive Data Vulnerabilities: Health data is sensitive personal information. Chatbot service providers will need to clarify how health data is stored, processed, shared, and used. Without strong data protection and transparency standards, users are exposed to further risks to their well-being. 

Way Forward

1. Addressing Therapeutic Misconception: A lack of understanding of the purpose and capabilities of such chatbots, in terms of care expectations and treatments they can offer, can jeopardize user health. Platforms providing such services should be mandated to lay disclaimers about the limitations of the therapeutic relationship between the platform and its users in a manner that is easy to understand. 
2. Improved Algorithm Design: Training data for these models must undertake regular updates and audits to enhance their accuracy, incorporate contextual socio-cultural factors for profile analysis, and use feedback loops from customers and mental health professionals. 
3. Human Oversight: Models of therapy where AI chatbots are used to supplement treatment instead of replacing human intervention can be explored. Such platforms must also provide escalation mechanisms in cases where human-intervention is sought or required. 

Conclusion

It is important to recognize that so far, there is no substitute for professional mental health services. Chatbots can help users gain awareness of their mental health condition and play an educational role in this regard, nudging them in the right direction, and provide assistance to both the practitioner and the client/patient. However, relying on this option to fill gaps in mental health services is not enough. Addressing this growing —and arguably already critical— global health crisis requires dedicated public funding to ensure comprehensive mental health support for all. 

Sources 

• https://www.who.int/news/item/17-06-2022-who-highlights-urgent-need-to-transform-mental-health-and-mental-health-care 
• https://health.economictimes.indiatimes.com/news/industry/mental-healthcare-in-india-building-a-strong-ecosystem-for-a-sound-mind/105395767#:~:text=Indian%20mental%20health%20market%20is,access%20to%20better%20quality%20services. 
• https://www.frontiersin.org/journals/digital-health/articles/10.3389/fdgth.2023.1278186/full 

‍

Overview:

It is worth stating that millions of Windows users around the world are facing the Blue Screen of Death (BSOD) problem that makes systems shutdown or restart. This has been attributed to a CrowdStrike update that was released recently and has impacted many organizations, financial institutions, and government agencies across the globe. Indian airlines have also reported disruptions on X (formerly Twitter), informing passengers about the issue.

Understanding Blue Screen of Death:

Blue Screen errors, also known as black screen errors or STOP code errors, can occur due to critical issues forcing Windows to shut down or restart. You may encounter messages like "Windows has been shut down to prevent damage to your computer." These errors can be caused by hardware or software problems.

Impact on Industries

Some of the large U. S. airlines such as American Airlines, Delta Airlines, and United Airlines had to issue ground stops because of communication problems. Also, several airports on Friday suffered a massive technical issue in check-in kiosks for IndiGo, Akasa Air, SpiceJet, and Air India Express.

The Widespread Issue

The issue seems widespread and is causing disruption across the board as Windows PCs are deployed at workplaces and other public entities like airlines, banks, and even media companies. It has been pointed out that Windows PCs use a special cybersecurity solution from a company called CrowdStrike that seems to be the culprit for this outage, affecting most Windows PC users out there.

Microsoft's Response

The issue was acknowledged by Microsoft and the mitigations are underway. The company in its verified X handle Microsoft 365 status has shared a series information on the latest outage and they are looking into the matter. The issue is under investigation.

In one of the posts from Microsoft Azure, it is mentioned that they have become aware of an issue affecting Virtual Machines (VMs) running Windows Client and Windows Server with the CrowdStrike Falcon agent installed. These VMs may encounter a bug check (BSOD) and become stuck in a restarting state. Their analysis indicates that this issue started approximately at 19:00 UTC on July 18th. They have provided recommendations as follows:

Restore from Backup:   In case customers have available backups prior to 19:00 UTC on July 18th, they should recover VM data from the backups. If the customer is using Azure Backup, they can get exact steps on how to restore VM data in the Azure portal. here.

Offline OS Disk Repair: Alternatively, customers can attempt offline repair of the OS disk by attaching an unmanaged disk to the affected VM. Encrypted disks may require additional steps to unlock before repair. Once attached, delete the following file:

Windows/System/System32/Drivers/CrowdStrike/C00000291*.sys 

After deletion, reattach the disk to the original VM.

Microsoft Azure is actively investigating additional mitigation options for affected customers. We will provide updates as we gather more information.

Resolving Blue Screen Errors in Windows

Windows 11 & Windows 10:

Blue Screen errors can stem from both hardware and software issues. If new hardware was added before the error, try removing it and restarting your PC. If restarting is difficult, start your PC in Safe Mode.

To Start in Safe Mode:

From Settings:

Open Settings > Update & Security > Recovery.

Under "Advanced startup," select Restart now.

After your PC restarts to the Choose an option screen, select Troubleshoot > Advanced options > Startup Settings > Restart.

After your PC restarts, you'll see a list of options. Select 4 or press F4 to start in Safe Mode. If you need to use the internet, select 5 or press F5 for Safe Mode with Networking.

From the Sign-in Screen:

Restart your PC. When you get to the sign-in screen, hold the Shift key down while you select Power > Restart.

After your PC restarts, follow the steps above.

From a Black or Blank Screen:

Press the power button to turn off your device, then turn it back on. Repeat this two more times.

After the third time, your device will start in the Windows Recovery Environment (WinRE).

From the Choose an option screen, follow the steps to enter Safe Mode.

Additional Help:

Windows Update: Ensure your system has the latest patches.

Blue Screen Troubleshooter: In Windows, open Get Help, type Troubleshoot BSOD error, and follow the guided walkthrough.

Online Troubleshooting: Visit Microsoft's support page and follow the recommendations under "Recommended Help."

If none of those steps help to resolve your Blue Screen error, please try the Blue Screen Troubleshooter in the Get Help app:

• In Windows, open Get Help.
• In the Get Help app, type Troubleshoot BSOD error.
• Follow the guided walkthrough in the Get Help app.

[Note: If you're not on a Windows device, you can run the Blue Screen Troubleshooter on your browser by going to Contact Microsoft Support and typing Troubleshoot BSOD error. Then follow the guided walkthrough under "Recommended Help."]

For detailed steps and further assistance, please refer to the Microsoft support portal or contact their support team.

CrowdStrike’s Response:

In the statement given by CrowdStrike, they have clearly mentioned it is not any cyberattack and their resources are working to fix the issue on Windows. Further, they have identified the deployment issue and fixed the same. Crowdstrike mentions about their problematic versions as follows:

• “Channel file "C-00000291*.sys" with timestamp of 0527 UTC or later is the reverted (good) version.
• Channel file "C-00000291*.sys" with timestamp of 0409 UTC is the problematic version.

Note: It is normal for multiple "C-00000291*.sys files to be present in the CrowdStrike directory - as long as one of the files in the folder has a timestamp of 0527 UTC or later, that will be the active content.”

The CrowdStrike will be providing latest updates on the same and advises their customers and organizations to contact their officials officially to get latest updates and accurate information. It is encouraged to refer to customer’s support portal for further help.

Stay safe and ensure regular backups to mitigate the impact of such issues.

References:

https://status.cloud.microsoft/

https://support.microsoft.com/en-us/windows/resolving-blue-screen-errors-in-windows-60b01860-58f2-be66-7516-5c45a66ae3c6

https://support.microsoft.com/en-gb/windows/start-your-pc-in-safe-mode-in-windows-92c27cff-db89-8644-1ce4-b3e5e56fe234

https://www.crowdstrike.com/blog/statement-on-falcon-content-update-for-windows-hosts/ 

‍

‍

‍

‍