It seems like everyone’s using AI, but I’ve heard it has security and environmental risks. Should I use it?
You’re right to be cautious. AI offers significant benefits, but it also presents legitimate concerns. The best way to mitigate those concerns, though, is to be equipped with the knowledge of how to use AI in a conscientious way, as we will review below.
Security First
Never share personal, confidential, or sensitive company information with AI tools. Many AI platforms retain and train on the data you provide, meaning your information might be stored and even reused. Remove any Personally Identifiable Information (PII) from materials before they are uploaded to an AI platform. Paid versions are less likely to retain data uploaded by users, but it is always good practice to check the terms of service for any tool before using it.
Accuracy
Always verify AI-generated outputs, especially if they include facts, statistics, or quotations; cross-reference them with trusted sources. AI tools can create outputs known as “hallucinations,” meaning AI responses that present themselves as facts but are false or non-existent. Additionally, if your query requires up-to-date information, be sure to check the knowledge cutoff date for the models training your platform of choice. For example, Google’s Gemini 2.5 Pro currently has a cutoff date of January 2025, and OpenAI’s ChatGPT-4-turbo has a knowledge cutoff of June 2024.
Bias Awareness
Biases in AI outputs can arise from a number of different factors. Most commonly, there may be historical biases and stereotypes in the training data of the AI models, which can lead to inadvertently biased content. However, even with otherwise neutral data, the algorithms and other design features of a particular platform could have been developed with inherent biases embedded in the system based on the humans creating or implementing it, which reinforces the need for representation and a variety of perspectives when producing any new technology. Furthermore, if an output contains bias or is interpreted in a biased manner and is then fed back into the system as new training data, this can create a feedback loop, amplifying biases. Always review AI outputs critically and consider the potential for biased responses.
Environmental Impact
AI’s rapid growth comes with substantial environmental costs. Data centers powering AI consume significant electricity and water resources. For instance, a single AI query can use up to ten times more energy than a traditional search, and training large models can emit as much carbon as five cars over their lifetimes. Moreover, AI data centers are projected to consume 4% of global electricity by 2030, up from 1–2% today. The good news is that there are some ways to minimize this impact:
- Choose Sustainable AI Tools: When possible, favor companies like Capalo AI and GridBeyond, that are transparent about their AI’s environmental impact and invest in energy-efficient infrastructure.
- Minimize Storage: Reuse and collaborate on files online instead of sending email attachments. Tools like MS Teams, which reduce email volume and file duplication, help cut digital carbon footprints. Also, regularly review and delete unnecessary documents to reduce cloud storage impact.
- Think Before You Search: AI-powered search engines use up to five times more energy than traditional search engines. Search smartly, and limit unnecessary queries. Use traditional search engines instead of AI chatbots for basic searches, and turn off AI features when they aren’t needed.
- Advocate for Better AI: Promote policies on carbon reporting and sustainability in AI development. Creating demand for this type of development and demonstrating public support for eco-friendly AI regulations helps elevate the cause and influences developers and policymakers alike.
Bottom Line:
AI can be incredibly helpful when used thoughtfully. By protecting your data, not taking outputs at face value, and being mindful of AI’s environmental footprint, you can harness its benefits without fueling its downsides. AI is here to stay, so, instead of ignoring it, learn how to use it properly and safely instead.
