AI alignment examines how artificial systems acquire, represent, and act on goals, values, and social norms, and why their behavior often diverges from human expectations. Work in this area studies systematic failures such as bias, sycophancy, hallucinations, deceptive or selfish reasoning, and cultural or linguistic inequities, as well as limitations in commonsense, emotion, and social understanding. It also develops methods for preference learning, norm-following, interpretability, and reliability guarantees to better align AI behavior with human values and societal constraints.
Software
The emergence of generative artificial intelligence (GenAI) models, including ChatGPT, Gemini and Copilot, has transformed educational pedagogy by enabling new forms of interaction, reflection and learning. Yet the use of ...
Jul 20, 2026
0
9
Machine learning & AI
When Eric Lauer needs to hire at Giftory, the online gift-giving platform he runs, he's no longer looking for eager young coders fresh out of college.
Jul 17, 2026
0
6
Machine learning & AI
Older people are being left out of decisions about how artificial intelligence is being built.
Jul 16, 2026
0
4
Machine learning & AI
Millions of people are now designing their own personalized artificial intelligence companions, yet most have little idea how those creations will actually behave. In a new paper, MIT Media Lab Assistant Professor Pat Pataranutaporn ...
Jul 16, 2026
0
3
Business
Can playing against AI make people more social? New research suggests the answer may be yes.
Jul 15, 2026
0
9
Machine learning & AI
Decision-making in clinical settings requires not only correct answers but also clinically valid reasoning. In a new study, researchers propose a framework to improve the use of artificial intelligence in medicine, making ...
Jul 14, 2026
0
7
Machine learning & AI
Early parent–child interactions lay the groundwork for communication, social and cognitive development. Yet many of the behaviors that signal how a child is developing are subtle, fleeting and difficult for untrained eyes ...
Jul 14, 2026
0
4
Computer Sciences
When can we trust the results we get from AI, and when is learning impossible? Researchers have shown that there are some problems that even the most powerful AI cannot reliably solve, no matter how much data it is given.
Jul 14, 2026
0
32
Machine learning & AI
Alan Turing, the father of theoretical computer science, made a proposal that has put AI development on a flawed path for three-quarters of a century, a prominent computer scientist has argued.
Jul 13, 2026
0
16
Consumer & Gadgets
Companies are increasingly using artificial intelligence in their hiring processes. It's not just CVs that are evaluated automatically. AI tools can also conduct job interviews—usually in the form of avatars, which are animated ...
Jul 10, 2026
0
11
