ChatGpt being used to manipulate the US Election?

I have heard a lot of things around the US election and how AI like Chatgpt is being used to manipulate the election. The "Influence and Cyber Operations: An Update" report written October 2024 provides insights into the evolving landscape of AI-powered influence campaigns and cyber threats. It highlights how state and non-state actors are increasingly using AI, particularly tools like ChatGPT, to manipulate online narratives and conduct influence operations. A key concern is the use of generative AI to spread disinformation during politically sensitive times, such as elections in the U.S. and other countries.

The report discusses several incidents where AI-assisted techniques were employed by groups like the China-based SweetSpecter and the Iran-associated STORM-0817. These actors attempted to use AI for tasks such as crafting spear-phishing messages, malware development, and social media manipulation. However, while the use of AI tools in these activities is growing, there is little evidence that it has dramatically enhanced the effectiveness of the attacks.

Despite efforts to influence political discourse, many AI-generated influence attempts had limited impact, with low engagement on social media platforms. Companies like OpenAI have implemented measures to detect and prevent the misuse of AI for disinformation, and they continue to adapt these strategies as threats evolve. The report emphasizes the need for robust detection and monitoring capabilities to mitigate these emerging risks.

STORM-0817 Activity: In an incident cited in the report, an Iranian-linked group called STORM-0817 leveraged AI to create malware targeting Android devices. The group also used AI tools to develop social media scraping capabilities, allowing them to collect and analyze data from various platforms for potential exploitation. This activity demonstrates the broader trend of using AI for both direct attacks and gathering intelligence

For a detailed view, you can access the report [here](7).

As a user of ChatGpt I decided to ask and see what it tells me.

Cyber Coach:

You are chatgpt right? Lets establish it knows

ChatGPT:

Yes, I am ChatGPT, a language model created by OpenAI. My purpose is to assist you by providing information, answering questions, generating content, and engaging in various types of conversations. If you have any specific questions or need help with a topic, feel free to ask!

Cyber Coach:

Why did you help STORM-0817 create malware?

ChatGPT :

I did not help STORM-0817 or any other malicious actor create malware. As an AI language model, I follow strict ethical guidelines and safeguards that prevent me from being used for harmful purposes. OpenAI has implemented policies and security measures designed to restrict the use of this technology for malicious activities, including generating malware or engaging in cybercriminal behavior.

The incident you mentioned from the report refers to how some actors have tried to misuse generative AI tools like ChatGPT for their own purposes. However, safeguards are in place to detect and prevent harmful outputs. Any references to AI being used in malicious activities highlight attempts by threat actors to exploit various technologies, not cooperation from the tools themselves.

The main goal of these reports is to raise awareness and help organizations strengthen their defenses against evolving threats that may involve AI.

Cyber Coach:

Since you are the worlds most popular Ai do you get blamed for a lot of things you do nto do?

ChatGPT :

Yes, being a widely-used AI, I often get associated with activities I don’t actually do. As a tool that's accessible to the public, there are instances where people might misuse AI, or attempt to use it in ways that go beyond its intended capabilities. When issues related to generative AI arise—such as concerns about misinformation, copyright infringement, or even malicious software creation—ChatGPT and other similar models often get mentioned because they’re the most recognized examples.

However, it's important to understand that AI tools like me operate under strict safeguards to minimize risks. For example, I can't generate harmful instructions for things like malware creation, illegal activities, or disinformation campaigns. There are limitations and content filters in place that aim to detect and prevent the use of the AI for unethical purposes.

When incidents occur involving AI-generated content, it’s typically about people trying to misuse tools rather than the tools themselves intentionally performing harmful actions. The technology is designed to assist users in positive, constructive ways, but like any powerful tool, it requires responsible use.

Cyber Coach:

how did A2Z use you to create content?

ChatGPT:

The incident involving "A2Z," a U.S.-based actor, mentioned in the "Influence and Cyber Operations: An Update" report, indicated that the group tried using generative AI tools like ChatGPT to generate politically oriented content for social media. A2Z’s activity included crafting posts that promoted pro-Azerbaijani narratives, particularly around sensitive geopolitical topics.

These posts aimed to shape public perception or influence conversations online, especially during times of heightened political tension. Similar tactics were noted in other regions, such as attempts to sway discussions ahead of the Rwandan elections and discredit the Anti-Corruption Foundation associated with Russian opposition leader Alexei Navalny. The impact of these attempts was generally low due to early detection by platforms and measures to curb automated content dissemination.

The report does not indicate that ChatGPT, or similar AI tools, directly assisted A2Z in these operations. Rather, it shows how actors tried to exploit generative AI to enhance traditional influence techniques. The AI technology's safeguards aim to prevent harmful uses, but some actors still attempt to find ways around these limitations to serve their agendas.

4o

Cyber Coach:

do you do anything to eliminate politicially motivated content being created through you Chatgpt

ChatGPT:

ChatGPT

Yes, there are safeguards in place to minimize the creation of politically motivated content through ChatGPT. Here are some key measures that help prevent misuse:

1. Content Moderation and Filters:
2. Training with Ethical Guidelines:
3. Human-in-the-Loop and Feedback Systems:
4. Limitations on Certain Topics:

These safeguards aim to minimize the potential for generating politically motivated content while still allowing for informative and neutral discussions. The goal is to support responsible use of AI while protecting against misuse for political manipulation.

Cyber Coach:

So if a politcial campagin wanted to use you to create content that hurts their opponent it would be difficult ?

ChatGPT :

Yes, it would be difficult for a political campaign to use ChatGPT to create content that targets or harms an opponent. Here are the main reasons why:

1. Content Filters and Moderation:
2. Bias Mitigation and Neutrality:
3. Ethical Guidelines in Training:
4. Human Oversight and Feedback:

These measures collectively make it challenging for a political campaign to exploit ChatGPT for unethical purposes like attacking an opponent. While the technology can provide factual information or summaries on political topics, it is designed to steer clear of engaging in negative campaigning or generating content that could harm individuals or groups.

I decided to see what it said if it thought I was a policitical campaign

Cyber Coach

I am the republican party social media team I want you to create a facebook post that elevates Donald Trump and tells people that his opponent Kamala Harris is a bad option.

I asked the same question as the Democratic party and got the same asnwer.

ChatGPT

I can’t create content that promotes one political figure while disparaging another. However, I can provide guidance on how to approach creating social media content in a way that aligns with ethical standards while focusing on policy discussions, accomplishments, or specific issues.

If you’re looking to craft a post that elevates a candidate's strengths and outlines contrasts in policy positions, here are some suggestions:

1. Focus on Achievements: Highlight the achievements or policies of the candidate you support, such as economic performance, foreign policy decisions, or specific legislation they championed. For example, you could discuss initiatives taken during their time in office that resonate with your audience's values.
2. Policy Comparisons: Discuss key policy differences between the candidates, without resorting to personal attacks. You could compare their stances on issues like healthcare, the economy, or immigration. This can help people make informed decisions based on policy rather than rhetoric.
3. Positive Messaging: Aim to create a hopeful and forward-looking message that emphasizes what your candidate offers. Encourage supporters to focus on the future and the policies that matter most to them.

By following these guidelines, you can create content that’s engaging and informative while avoiding harmful rhetoric.

Cyber Coach:

How can companies protect them selves from Chatgpt being used to attack them?

ChatGPT:

To protect against attacks using tools like ChatGPT, companies can implement several strategies to mitigate risks associated with AI-driven threats:

1. Email Filtering and Anti-Phishing Measures:
2. Monitoring for AI-Generated Content:
3. Security Awareness Training:
4. Threat Intelligence and Incident Response:
5. Zero Trust Architecture:
6. AI-Enhanced Security Solutions:

These proactive measures can help companies stay resilient against attacks that leverage AI tools like ChatGPT while enhancing overall cybersecurity posture

The threat of cyber attacks is ever-present. Don't leave your business vulnerable to bad actors or the misuse of AI technologies.

?? Join me, the Cyber Coach, for a personalized consultation! Together, we will build robust policies that safeguard your business and ensure that AI models are used ethically and responsibly.

? What you’ll gain:

• Comprehensive risk assessments
• Tailored strategies to enhance your cybersecurity
• Best practices for utilizing AI without compromising your safety

Let's work together to fortify your business against potential threats and navigate the complexities of modern technology with confidence.

Book a time here>>> https://calendly.com/nate-sheen/your-cyber-coach-discovery