Please enable JavaScript.
Coggle requires JavaScript to display documents.
AI and ML - Coggle Diagram
AI and ML
Cybersecurity Advantages
Enhanced Threat Detection
can recognize and alert users to unidentified malware, shady activity, and network breaches
Automated Security Operations
Real time Incident response
Adaptive Defense Mechanisms
Reduced False Positives
reduces warning fatigue
Privacy and Security Challenges
You realize sensitive information can be decrypted very easily by an adversarial attack to an AI model you sold to a client.
Update the client immediately. Correct the error to include encryption and deliver the updated product to the client.
This is the ethical choice due to the substantial amount of damage that could be done to the client and the owners of the sensitive information. Utilitarianism
Let the client worry about their own security.
An artist believes the AI I'm working on is using their art to generate similar images.
Use XAI to break down complex ML models into easily understood verifiable components.
This is the best option for everyone because there are little to no consequences, and it is fair to all stakeholders. It also shows virtuous character.
Don't respond because you know what training data you used.
Cybersecurity Risks
You believe the training data has been poisoned, and the model has become biased.
Find new training data or find an uncorrupted version of the training data.
This will create the least amount of harm from the decisions made by your model. It also ensures everyone under the data set will be treated in the same manner.
Keep the current training data.
You believe a complex AI phishing campaign is happening at your workplace.
Educated the rest of the staff of your suspicions.
This create the most amount good for the company and the least amount of harm to the employees.
Mark the email as spam and continue your day.
Ethical Challenges
Job displacement
A new AI will save the company money if we replace some of the workforce for the AI.
Keep workers.
Replace the workforce.
Replace workers while offering support to retrain in another role.
This is the most ethical option because it balances the need for technological advancement and minimizes the consequences for the workers. By trying to keep the workers it show virtuous business culture and respects the dignity and autonomy of the workers. I believe this is the best outcome for all stakeholders involved to keep the business sustainable.
Prejudice
An applicant is rejected a job even though you know they are qualified.
Take the time to ensure training data's size quality, and representativeness will create a non-biased hiring model. Evaluate and validate model performance.
This is the most ethical decisions because it ensures Universalistic principles of everyone being treated the same in similar situations.
Trust the AI has made the right decision.