ARTIFICIAL intelligence (AI) deserves criticism. Its environmental costs, embedded biases, labor implications, capacity to manufacture falsehoods and possible reliance on copyrighted material require serious scrutiny. The corporations developing AI must be held accountable, and users must observe ethical boundaries.
But criticism is one thing. Outright rejection is another.
Some critics treat any use of AI as inherently unethical. For them, it does not matter whether the output is accurate, disclosed, reviewed or responsibly used.
The mere involvement of AI supposedly contaminates the product. A graphic assisted by generative AI is dismissed as theft. An essay improved using an AI editing tool is called fraudulent. A presentation containing an AI-generated illustration becomes morally suspect regardless of its educational value.
This is not ethical discernment. It is technological absolutism. Ethics requires context and judgment, not slogans masquerading as moral certainty. And moral consistency requires critics to acknowledge the AI embedded in the very technologies through which they proclaim their opposition to it.
It also becomes intellectually awkward when the same critics use smartphones, social media, streaming services, search engines, navigation applications, digital cameras, online shopping sites, grammar checkers and automated design tools. These technologies either contain AI systems, depend upon AI-enabled infrastructure, or use algorithms performing functions within the broad territory of artificial intelligence.
Their phones use computational photography to improve images. Their email providers filter spam. Social media platforms rank their feeds algorithmically. Navigation applications calculate routes using predictive models. Streaming services recommend content. Search engines interpret queries. Accessibility tools convert speech into text, produce captions and recognize images. Even ordinary editing software now uses object detection, image enhancement and language prediction.
One may distinguish generative AI from older machine-learning applications, and that distinction matters. Generative systems pose specific questions about training data, authorship, fabrication and creative labor. But this does not justify the sweeping claim that every use of generative AI is necessarily unethical. Technologies should be judged according to what they do, how they were developed, how they are used, who benefits, who is harmed and what safeguards govern them.
The ethical question is not simply: Was AI used?
It should be: For what purpose was it used, under what conditions, with what degree of human oversight and with what consequences?
Consider education. A professor may use AI to create a simple conceptual illustration for a lecture on climate resilience. The image is not sold as original art, entered in an art competition, or represented as the professor’s hand-drawn work. It merely helps students understand a complex relationship. If the professor checks it for accuracy, discloses its origin when relevant, and avoids imitating a living artist’s distinctive style, it is difficult to argue that this use is inherently immoral.
AI can also help teachers produce alternative explanations for students with different levels of preparation. It can simplify overly technical prose, propose discussion questions, translate instructional material, or generate a preliminary diagram that the teacher subsequently corrects. These applications do not eliminate the teacher’s responsibility. They require the teacher to exercise judgment more actively.
The same applies to writing. Using AI to fabricate evidence, invent citations, impersonate someone, or generate a paper that one submits as entirely one’s own is plainly irresponsible. But using it to detect grammatical errors, test alternative structures, summarize one’s own notes, or suggest clearer wording is not fundamentally different from working with a human editor, provided the author remains responsible for every argument, fact and sentence in the final work.
Responsible use is also possible in other fields. Doctors may use AI-assisted systems to detect abnormalities in medical images, but clinical decisions must remain with qualified professionals. Disaster managers may use predictive tools to analyze rainfall, flooding or evacuation requirements, but the models must be validated. Researchers may use AI to organize large bodies of literature, but they must verify every citation. Government agencies may automate routine work, but decisions affecting rights and benefits must remain explainable, reviewable and appealable.
Journalists can use AI to transcribe interviews while checking the output against the recording. Persons with disabilities can use image descriptions, speech recognition, predictive text and real-time captions. Small organizations with limited resources can use AI to develop initial layouts or visualize ideas before seeking professional assistance for final production.
These are not arguments for unrestricted adoption. They are arguments for discrimination in the best sense of the word: the capacity to distinguish responsible uses from irresponsible ones.
Responsible AI requires clear norms. Users must disclose material AI involvement when authorship, assessment, authenticity or public trust is at stake. AI-generated facts and citations must be independently verified. Confidential information must not be casually uploaded into public systems. Organizations must examine automated decisions for bias. Artists must be protected from unauthorized imitation and deceptive substitution. Workers should not simply be discarded in the name of efficiency. Human beings must remain accountable for consequential outputs.
Most importantly, AI should augment human judgment, not become an excuse for abandoning it.
The proper response to a technology with benefits and dangers is governance, not ritual denunciation. We did not ban calculators because students could use them to avoid learning arithmetic. We developed rules concerning when they could be used. We did not reject the internet because it enabled plagiarism and disinformation. We developed digital-literacy programs, imposed standards and continued debating platform regulation. AI demands the same mature approach.
There is nothing wrong with personally refusing to use generative AI. Artists may choose entirely human-made processes. Teachers may prohibit it in certain assignments. Writers may work without automated assistance. These choices deserve respect.
What does not deserve automatic respect is turning personal preference into universal moral law, particularly when preached through an AI-ranked social media platform, typed on an AI-enabled smartphone, corrected by predictive text, enhanced by computational photography and distributed by algorithms designed to maximize its visibility.
Critique AI rigorously. Regulate it firmly. Use it transparently and responsibly. But spare us the performance of technological purity.
Condemning every form of AI through a gadget and platform powered by AI is not resistance. It is merely hypocrisy with autocorrect.
The author is a professor at the University of the Philippines Los Baños and vice chairman of the board of People’s Television Network Inc.