November 2025 - Vol. 68 No. 11

While Silicon Valley is investing tens of billions of dollars chasing the AGI dream, academic computing research in the U.S. is facing a severe drought.

We look forward to working with authors whose articles expand both the utility and the delights of computing.

A career is not about success but about succeeding on your own terms.

A look at the mechanics and economics of the AI-enabled cybercriminal frontier, plus actionable insights for defense.

Scientists are using machine learning techniques such as classifiers to identify units of language and hope deep neural networks will help them figure out what animals might be talking about.

AI-assisted renderings can show historic structures in precise detail across different time periods.

LLMs and generative AI are rapidly transforming how Web searches are conducted and delivered.

Software researchers can make their research more accessible by actively approaching those communities that stand to benefit most from their findings.

A common complaint is that developers infringe by using in-copyright materials as training data for the foundation models that power their AI systems.

Human users are prone to believe that a generative language model is erudite and intelligent, conflating linguistic competence with other human characteristics.

Digital trade is a new engine of economic growth that alters traditional notions of trade balances.

Breaking the dogma that AI requires huge computing power and huge datasets could open up new technological horizons.

Tech workers have historically been monumentally uninterested in unionization, but now is the time to get organized.

There are still ways to further reduce the overhead of programs that span the JavaScript/Wasm boundary.

Identifying winning and losing patterns of Wasm deployments.

Static analysis is helping the research community improve automatic C-to-Rust translators.

The likelihood of AI serving as an omnipotent, truth-saying oracle soon is not great.

An overview of cyberpsychology’s contributions to the field of computing and technology.

Shielding is a principled and effective approach to ensuring the safety of reinforcement learning, even during training.

Cryptocurrency-themed browser extensions offer utility and convenience but also new opportunities for cybercriminals to steal cryptocurrency, launch phishing attacks, or conduct unauthorized crypto mining.

This work characterizes cryptocurrency-themed malicious extensions and encourages extension store operators to enact dedicated countermeasures.

A young man's desperation to recount a forgotten part of his memory leads him down a dark path.