5 Risks of Artificial Intelligence

Everyone will have an AI-powered personal assistant within 5 years- a prediction by Bill Gates. The Microsoft co-founder thinks the potential rewards outweigh the big risks that come with AI. He called out five big risks with the technology in the future despite predicting that AI's "benefits will be massive.

1. Deepfakes: Deepfakes are manipulating facial appearance through deep generative methods. While the act of creating fake content is not new, deepfakes leverage powerful techniques from machine learning and artificial intelligence to manipulate or generate visual and audio content that can more easily deceive.

The sophistication of AI deepfakes is growing daily, and the information we disclose online is feeding the crooks. They can pretend to be loved ones by using technology.

"People can use AI to replicate someone else for a variety of reasons, even with limited tech knowledge.

False sounds that resemble friends and family in distress might be used by scammers. They extort money using the deepfakes. Additionally, it's easier the more details they know about you.

Since it is simple to mimic the voices of those who post their voices online, the only way to be sure is to end the call, give the family member a callback

Verify their location and contact local law enforcement right away. Create a password with family and friends that you can share in the event of a scam. Limit the information you share online, particularly if you’re traveling. It’s like putting up a sign outside your home saying, ‘We’re not home.’ We wouldn’t do that in real life, but we feel comfortable doing it on social media.

2. Helping hackers: Security researchers have shown that generative AI is capable of social engineering humans, and AI algorithms can be used to collate vast amounts of personal information about individuals, assisting hackers in crafting highly convincing spear phishing emails.

Hackers are turning to artificial intelligence and using it as a weapon for malware and attacks to counter advances in cybersecurity solutions. For example, criminals use AI to hide malicious code inside benign applications. They program the codes to run at a specific point in time, such as ten months after the programs are installed or when a certain number of users have subscribed to the programs. This is to maximize the impact of such attacks. Hiding such code and data requires applying artificial intelligence models and deriving private keys to control where and when the malware is executed. However, hackers can configure the app's functionality as an AI trigger to carry out cyber attacks. Features can range from authentication processes to voice or visual recognition to identity management capabilities. Most applications in use today include such features, and this gives attackers many opportunities to feed weaponized AI models, obtain keys, and attack if desired. Malicious patterns can exist for years without being detected because hackers wait to attack when applications are most vulnerable.

Additionally, AI technologies are unique in that they acquire knowledge and intelligence to adapt accordingly. Hackers are aware of these features and use them to model adaptive attacks and create intelligent malware programs. Therefore, during attacks, programs can gather information about what prevented the attacks from succeeding and retain information that is useful. AI-based attacks may not succeed on the first try, but adaptive capabilities can help hacker succeed in subsequent attacks. Therefore, the security community must gain a deep understanding of the techniques used to develop AI-based attacks in order to create effective mitigations and controls.

Cyber adversaries also use artificial intelligence to conduct intelligent attacks that propagate through a system or network. Intelligent malware can exploit an unlimited number of vulnerabilities, increasing the likelihood of fully compromised targets. When a smart attack encounters a patched vulnerability, it immediately adapts to try to compromise the system with different types of attacks.

3. Eliminating jobs: One of the greatest disadvantages of AI is that it will eliminate jobs. Bill Gates acknowledged that AI will replace human workers in some scenarios.

4. Biases and hallucinations: AI apps sometimes "hallucinate" by making up things that aren't genuine. Another related issue is that AI can reflect biases included in their training data. Gates noted that AI could even make existing biases worse.

5. Student cheating: It is feasible for students to turn in work to their teachers that AI actually did. Gates acknowledged that such cheating is a valid problem.

In the 1970s and 1980s, electronic calculators became widely used Some math teachers were worried their students wouldn't learn how to do arithmetic without a calculator and gates compared the current situation to the past.