Ads generated by artificial intelligence are nearly indistinguishable from human-made ones, but new research shows they consistently underperform compared to human-made work when it comes to predicting short-term sales impact.

罢丑别听 from global research firm Ipsos in collaboration with two faculty members from the聽 tested 20 ads across 10 brands with 3,000 U.S. respondents. They found that human-made ads outperformed their AI counterparts, though the gap between the two was surprisingly slim.

The study paired existing human-made ads, produced before 2021 to ensure AI tools were not used, with fully AI-generated counterparts built from the same strategic brief, the document that ad professionals use to outline objectives, messaging and tactics for a campaign. Ads were then viewed by real consumers.

The results challenge assumptions the advertising industry can no longer afford to ignore, faculty 听补苍诲听 say, while the project overall reflects Newhouse鈥檚 commitment to train students with the skills and forward-thinking strategies needed to be effective and ethical communicators.

The Research Team

Peruta, director of the聽聽M.S. program, and Riby, professor of practice in the聽, led the University side of the study. Ryan Barthelmes, senior vice president of creative excellence at Ipsos, guided the project for the research firm.

Peruta oversaw the technical process of deconstructing existing ads and building the pipeline to produce their AI counterparts. AI was assigned to do everything a creative team would do, from interpreting strategy to developing a concept to producing the final spot.

鈥淭he human ads and the AI ads started from the same brief,鈥� Peruta says. 鈥淭he only thing that changed was who made them, and that鈥檚 exactly what we wanted to measure.鈥�

Riby brought advertising strategy and creative expertise, including insights drawn from her The Big Idea in Advertising class, where Newhouse students have spent three years creating AI-generated ads and evaluating the results.

The 10 brands selected for the project spanned various sectors, including consumer packaged goods, fashion, automotive and technology: Cheerios, Chewy, Febreze, Fiat, H&M, Old Navy, Herbal Essences, Ray-Ban Meta, TurboTax and Visa.

Raina Rice 鈥�26, an advertising major, supported the project behind the scenes, helping organize and manage the ad assets across all 10 brand pairings.

What They Found

The study produced three findings that promise to generate conversation across the advertising industry.

• Consumers largely cannot tell the difference.聽Only 13% of viewers who saw an AI-generated ad were at least somewhat confident it was made by AI鈥攖he same share as viewers who suspected human-made ads were AI-generated. With 40% of all viewers uncertain either way, the line between human and machine-made advertising is blurring quickly.

• Despite that perceptual similarity, a measurable effectiveness gap emerged.聽Using Ipsos鈥� sales-validated measures of advertising performance, human-made ads over-indexed against the benchmark by 11 points on average, while AI-made ads under-indexed by five. In practical terms, human ads are predicted to drive stronger short-term sales impact. AI can produce credible work, but on average it does not move the needle the same way.

• AI performed best when the brief was straightforward and product-driven, but struggled when the creative challenge called for storytelling, emotion or a genuine point of view.聽The strongest result in the study came from the Cheerios pairing, where a deeply human brief produced the highest combined effectiveness scores across both versions.

鈥淓very semester in my class, I watch students create AI ads about themselves, and not one of them has ever loved their output enough to put it on their refrigerator,鈥� Riby says. 鈥淭hat reaction is the premise of this entire study. If the creators themselves are underwhelmed, why would we expect consumers to feel differently? The data now backs that up.鈥�

An Industry Perspective

Barthelmes says the study addresses a question the advertising industry has been circling but is reluctant to answer directly.

鈥淓very [chief marketing officer] is being asked whether AI can replace their creative agencies, and creative directors are wondering about their futures,鈥� Barthelmes says. 鈥淭his research gives us a framework for that conversation. AI is a powerful tool, but the data shows that the human capacity for storytelling and emotional connection still creates a measurable competitive edge. The future is humans and AI working together.鈥�

Looking Ahead

The Newhouse-Ipsos partnership reflects the school鈥檚 broader investment in industry-facing research that shapes how the next generation of communicators understands and works alongside AI.

The study鈥檚 key recommendation is clear: do not settle for 鈥済ood enough.鈥� AI has an important role in modern campaign strategy and execution, but it is not a replacement for the human-led creativity needed to deliver a competitive advantage.

Artificial intelligence (AI) is already making substantial changes in every industry, shifting how we work, learn and organize our daily lives. But how can AI tools shape the field of building science? That was the central question at the Industry Summit on Artificial Intelligence for the Built Environment, organized by , Traugott Professor of Mechanical and Aerospace Engineering in the College of Engineering and Computer Science and co-director of the (性视界CoE).

Structured as a working session, the May 4 summit featured expert panelists from industry, academia and government agencies, with 12 companies represented and a total of 35 participants. After opening remarks from Professor Dong, the first panel of the day explored AI applications in smart and human-centered buildings. Presentations included:

• From Equipment to Ecosystem: An AI Strategy for Thermal Energy Systems and the Built Environment, presented by Josiah Johnston, senior director of data science at Daikin Open Innovation Lab Silicon Valley
• AI in Buildings: A Perspective From the Field, presented by William Healy, senior director at TRC Companies
• Using AI for Building Optimization, presented by Evan Torkos, vice president for strategy at Nantum AI
• The Restoration of a Building or Home鈥檚 Comfort, a New Set of Opportunities With AI, presented by Michael Birnkrant, chief architect, service and aftermarket at Carrier Corporation

A moderated discussion led by 性视界CoE鈥檚 executive director, , gave attendees a chance to dig deeper into these AI advances before breaking for a student poster session and lunch.

The afternoon panel widened the lens to AI鈥檚 role in building-connected infrastructure, covering the following topics:

• Load Flexibility and Electrified Commercial Buildings, presented by Mark Bremer and Julia Griffith from National Grid
• Hallucination of AI in Critical Infrastructure, presented by Herbert Dwyer, founder and CEO of EMPEQ
• A Semantic Foundation Unlocks Rapid Deployment of AI in the Built Environment, presented by Andrew Rodgers, co-founder of ACE IoT Solutions
• AI-Powered Communities: From Data to Resilience, presented by Nancy Min, co-founder and CEO of ecoLong
• Using GenAI to Accelerate Decarbonizing NYC Commercial Real Estate, presented by Thomas Yeh, consulting technical advisor, NYSERDA

The summit concluded with small group discussions: four breakout groups each co-facilitated by 性视界 University faculty and populated with a cross-section of academic and industry voices. This format ensured that the day鈥檚 themes were stress-tested in conversation and built the foundation for future collaborations. Dong plans to apply for funding for an interdisciplinary research center, such as a National Science Foundation Engineering Research Center, that will advance university-industry partnerships in the healthy buildings field.

The summit made clear that AI鈥檚 role in the built environment is no longer speculative鈥攊t is operational and growing rapidly. From smarter HVAC to grid-scale flexibility to community resilience, the challenge now is deploying these tools thoughtfully, sustainably and at scale.

This event was supported by the University鈥檚 聽through their Team Building for Large, Collaborative Grants program.

To be notified of future events and opportunities, sign up for 性视界CoE’s 听辞谤听.

What happens when students are allowed to use artificial intelligence to solve cybersecurity challenges? That question took center stage as (EECS) Professor hosted a capture-the-flag (CTF) cybersecurity competition at the , bringing together 20 undergraduate, master鈥檚, and Ph.D. students.

Unlike traditional CTF competitions, participants in this event were allowed to use modern AI assistants, such as ChatGPT and Claude, while solving challenges. The competition was designed not only to test technical skills, but also to explore how AI is transforming the way students learn and approach complex cybersecurity problems.

鈥淐ybersecurity education is evolving rapidly with the rise of AI tools,鈥� Hoque says. 鈥淭his competition gave us a unique opportunity to observe how students use AI in real time鈥攚hether it helps them think more deeply about problems or simply speeds up solutions. Understanding that distinction is critical for the future of computer science discipline.鈥�

Participants competed individually across 10 challenges spanning beginner, intermediate and advanced levels. The top three performers鈥擶eixiang Wang (first place), Annepu Sai Charan (second place) and Armani Isonguyo (third place)鈥攚ere ranked based on the number of challenges solved and the speed at which they completed them. Students described the experience as both exciting and challenging, noting that AI could guide their thinking but still required careful verification.

鈥淭his reflects how we approach computer science and cybersecurity education at 性视界 University,鈥� says Alex Jones, the Klaus Schroder Professor and chair of Electrical Engineering and Computer Sciences. 鈥淎I tools are only as effective as their operators. They do not replace expertise. Dr. Hoque鈥檚 work is a great illustration of this approach. We emphasize deep fundamental knowledge while also encouraging the use of AI. This ensures our graduates can effectively use, evaluate, guide, and validate AI-driven solutions.鈥�

To better understand the educational impact of AI-assisted problem solving, Hoque collaborated with Farzana Rahman, an expert in computing and AI education. Together, they are investigating how students use AI tools, whether those tools support meaningful learning and how they influence confidence and problem-solving strategies.

鈥淲e鈥檙e seeing a fundamental shift in how students engage with complex technical tasks,鈥� says Rahman. 鈥淎I can be a powerful learning aid, but we need to understand how to use it without compromising deep technical learning.”

Hoque plans to expand the CTF initiative by offering additional training sessions and forming student teams for regional and national competitions, further strengthening cybersecurity engagement within the EECS community.

The event is part of Hoque鈥檚 broader efforts, including , to advance education at the intersection of cybersecurity and artificial intelligence.

Professor has developed an artificial intelligence bot that uses curated, course-specific materials to assist students in mastering the applicable legal rules and their application. These digital 鈥渃oaches鈥� are available 24/7 to assist students in understanding challenging concepts and then to quiz students on their application, providing immediate feedback in a variety of question and answer formats. Thus far, Graves has deployed the concept in his evidence and contracts courses.

Graves uses OpenAI鈥檚 private custom GPT feature, which allows him to provide students with an interactive experience that is narrowly tailored to his specific course. Graves accomplishes this with a comprehensive set of instructions (i.e., prompts) telling the custom coach exactly what to do鈥攑roviding guardrails to keep it focused on the objective of assisting students in this course鈥攁nd uploading copies of the course text and other key instructional materials that facilitate Retrieval Augmented Generation (RAG). This domain-specific RAG layer increases the accuracy of the coach鈥檚 responses in the context of this specific course and dramatically reduces the potential for errors, as compared to a generic Large Language Model (LLM) trained on generic data of varying quality.

This 鈥渨alled garden鈥� of course-specific material addresses the common issue with LLM AI platforms that indiscriminately draw from all information on the internet.

鈥淭he LLMs pick up a good deal of erroneous information from unreliable sources, and they miss a lot of really good information that鈥檚 behind firewalls,鈥� Graves says. 鈥淭he bot has been instructed to respond to students when they ask for answers by walking them through in a Socratic-style dialog much as I might in class or office hours. When assisting students, the coach relies first and foremost on the information uploaded in its RAG layer, not only helping to explain and quiz the students on accurate course doctrine, but pointing students directly to appropriate sources within the course text itself.鈥�

Of course, the key to this approach is a collaborative relationship between Graves and the publisher of his course textbooks. While Graves is a co-author of his contracts textbook, the copyright is held by West Academic (the publisher of both the Learning Evidence and Learning Contracts textbooks used by Graves).聽 Graves worked closely with West Academic in developing an approach that would appropriately protect all copyrighted material uploaded to the coach鈥檚 RAG layer, and his use of both Learning Evidence and Learning Contracts is done under license from West Academic.

The use of the primary course text within the RAG layer effectively expands the value of the text far beyond the initially assigned readings. At the core of the coach鈥檚 domain-specific content, the textbook continues to anchor the coach鈥檚 role in assisting and quizzing students as they better learn to apply that content.

Supplementing the Teacher’s Role

Graves says, 鈥渢he teacher鈥檚 role is not being outsourced to the coach鈥攊t is being supplemented in new ways for which narrowly tailored AI is uniquely suited.鈥�

鈥淭he Coach does not replace basic course prep or attendance,” he says. “It is purely a supplement to these traditional teaching and learning tools鈥攁lbeit a very effective one, arguably far more effective than traditional generic study aids or generic LLMs often used by students today. Perhaps most valuable is the coach鈥檚 ability to provide students with unlimited opportunities to apply the course material in a variety of assessment formats, all of which are subject to immediate feedback. At the end of the day, this is often the single most effective teaching and learning tool for law students, and the coach provides this tool in a manner that is always available and fully aligned with course content and course outcomes.”

Students access the coach through a dedicated course link, which provides for private interaction between student and coach, unless the student voluntarily decides to share the unique link generated by a specific conversation. The initial privacy of the conversation encourages students to ask questions they might otherwise be uncomfortable raising (the proverbial 鈥渄umb question,鈥� which is often anything but).

It also allows students to use the coach in collaborative study sessions or to forward a conversation to Graves for further exploration. This latter feature is particularly useful in terms of quality control of both student prompts and responses by the coach.

鈥淒uring the past two semesters, I鈥檝e seen a few responses from the coach that could be improved and one blatant error,” Grave says. “However, the vast majority of interactive challenges arose from imperfect student prompts.鈥�

Thus, the students get two additional benefits from using the coach: they learn the importance of effective inputs (prompts) and they learn the importance of verifying outputs.

Continuing to Fine-Tune the Tool

While the evidence and contract coaches have proven very accurate (Graves directly tests them regularly himself, in addition to frequent student feedback), AI remains imperfect, and the professor has continued to 鈥渇ine-tune鈥� his bots by uploading additional course-specific material based on his own testing and observations of student/coach interactions.

Graves teaches exclusively in the College of Law鈥檚 , so the 24/7 availability of his coaches is particularly important to a body of students located around the world.

鈥淭his has allowed me to be more efficient and effective with my time while giving our global students a uniquely tailored experience that will help them master course material, while being available at any time that is convenient to them,鈥� he says.

Ask what keeps him up at night about artificial intelligence and you won鈥檛 get a single answer.

The University鈥檚 senior vice president for digital transformation and chief digital officer is tracking several threads at once: how AI can reshape higher education, why the job market isn鈥檛 collapsing the way headlines suggest, what it will take to rebuild trust in online content, the need for regulation and where the University鈥檚 massive stores of data fit into all of it.

Rubin shared some of his recent thinking as a panelist at a Maxwell School fireside chat on digital transformation and AI in New York state. Here are four takeaways.

Artificial intelligence (AI) is reshaping how governments operate, how universities teach and how public institutions make decisions.

That was the central message of a recent fireside chat hosted by the . Dean moderated the conversation which brought together two leaders working at the forefront of AI adoption: , commissioner of the New York State Office of General Services (OGS), and , 性视界 University鈥檚 senior vice president for digital transformation and chief digital officer.

鈥淭he question before us is not whether AI will transform public life,鈥� Van Slyke said. 鈥淚t鈥檚 whether our institutions are ready to lead that transformation thoughtfully, equitably and effectively.鈥�

Personalizing Learning and Expanding Access

Rubin opened the March 26 event with a claim about the stakes for higher education: AI, he said, has the potential to transform how universities teach in ways not seen in 200 years. 鈥淭he idea of a professor standing in front of a room, lecturing鈥攁nd students taking notes and then being assessed through projects, papers and exams鈥攖hat model has not shifted,鈥� he said. 鈥淲hat AI allows you to do is personalize learning.鈥�

Personalization at scale has long been a challenge because no instructor can simultaneously tailor a course to every student鈥檚 pace and needs, he said. AI changes that equation.

Rubin shared how 性视界 has deployed more than 30,000 AI licenses across campus to drive equitable access and data security. Some students had already purchased AI tools on their own, while others could not afford them, he pointed out. Faculty and staff also needed a secure environment for uploading sensitive documents without routing data through commercial platforms.

Rubin also highlighted a less-discussed dimension of the University鈥檚 AI work: a private wireless network, built in partnership with JMA Wireless, that supports thermal sensors in academic buildings across campus. The sensors detect occupancy without capturing identifying information, allowing the University to optimize janitorial services, plan building capacity and, eventually, adjust heating and cooling based on actual use patterns.

A Measured Approach to Government AI

Moy noted that the state鈥檚 deliberate pace of technology adoption is a necessary safeguard rather than a liability. 鈥淚 would contend that it鈥檚 important that government is risk-averse,鈥� she said. 鈥淭he information that we hold is really important鈥擬edicaid data, health data, testing information. The importance of that stewardship becomes paramount.鈥�

Her office oversees roughly 30 million square feet of state real estate, manages 1,500 procurement contracts valued at $44 billion and administers a design and construction portfolio of approximately $5.7 billion. Moy described the agency鈥檚 AI strategy as a measured approach. It involves first identifying low-risk, high-value applications, then building the data infrastructure to support them, and ensuring legal and operational frameworks are in place before scaling.

Moy said one of OGS鈥檚 most tangible AI investments is in procurement search. Agencies and municipalities navigating the state鈥檚 contract catalog often struggle to find what they need, undermining the efficiency those contracts are designed to provide. Moy said AI-assisted search is a logical starting point: low risk, no job displacement and an immediate opportunity to test what the technology can do.

The agency is also piloting AI-powered document summarization tools for bid documents and contract histories which are reported to save up to three hours per day.

Moy noted that backlogs present another opportunity, as they are a universal challenge across the public sector. She explained that while AI could help alleviate some of those challenges, agencies must be cautious; they cannot hand out productivity tools to every worker without first creating the right frameworks.

Jobs, Regulation and What Comes Next

Both speakers addressed audience concerns about AI鈥檚 impact on jobs鈥攁 topic that has gained urgency in New York following Governor Kathy Hochul鈥檚 , which is tasked with studying AI鈥檚 effects on the labor market.

Rubin cited research suggesting that less than 1% of the 1.2 million layoffs recorded in 2025 were directly attributable to AI, arguing that economic factors and structural business decisions are doing more to reshape the workforce than the technology itself. He expressed confidence that AI will ultimately create more jobs than it displaces, though he acknowledged that every job will change.

鈥淚f you don’t know how to incorporate AI into your domain and discipline, you will be at a disadvantage,鈥� he said. 鈥淪tudents need to have the tools and the classes.鈥�

Moy recalled the dot-com era and the transformation of publishing that upended models at institutions like the Brooklyn Public Library, where she once served as chief strategy officer. The fear and exuberance that accompanied those transitions, she said, mirrors what society is experiencing today.

鈥淲e want to make sure that we鈥檙e thinking about it ethically, that we鈥檙e balancing it according to public need,鈥� she said. 鈥淎nd we鈥檙e having active conversations about those trade-offs.鈥�

Both panelists returned repeatedly to the theme of transparency in AI systems, government data and institutional communications.

Rubin pointed to Anthropic鈥檚 practice of publishing system prompts as a model for responsible AI deployment and noted that 性视界 recently launched an AI-powered course search tool, called , that similarly makes its operating parameters visible. He also raised the challenge of AI-generated media and the difficulty of distinguishing real content from fabricated content online.

An Open and Ongoing Dialogue

The conversation drew questions from the audience.

A first-year Maxwell student and member of the University’s United AI club asked what precedent a recent court ruling holding social media platforms liable for algorithmic harm to minors sets for the future of AI regulation and whether platforms like ChatGPT should face similar oversight.

Rubin was direct: 鈥淲e made the mistake with social media. These companies should have an obligation to have guardrails.鈥�

Moy pointed to Hochul鈥檚 recent policy proposals targeting addictive technology, including requirements for more restrictive default settings on children鈥檚 accounts. She acknowledged that government is often a step behind rapid technological change, but argued that intervention becomes necessary when innovation results in public harm.

A second student raised concerns about AI鈥檚 potential to enable fraud, including falsified documents and biased algorithms.

鈥淭hese are very real questions,鈥� she said, emphasizing that OGS is working to understand its uses and risks. She argued that the answer isn鈥檛 avoiding AI but understanding it well enough to spot its misuse. 鈥淚f we don’t understand it, we will fall behind.鈥�

Rubin agreed, framing the detection challenge as both technological and philosophical: As AI becomes embedded in everything from autocomplete to document editing, defining what counts as 鈥淎I-generated鈥� becomes increasingly difficult. 鈥淢y gut is almost every piece of content out there will have some AI piece to it, assisting us,鈥� he said. 鈥淪o, it鈥檚 a technology challenge and a societal challenge.鈥�

Van Slyke closed by noting that Maxwell鈥檚 role in preparing students for public service has always meant equipping them not just with technical knowledge, but with the ability to navigate the policy, governance and ethical dimensions that accompany it.

鈥淭he question is not what will AI do to our institutions,鈥� he said. 鈥淚t’s what will we choose to do with it.鈥�

When Tyler Neary 鈥�27and Orion Goodman 鈥�27 scattered flyers across campus last spring advertising a new AI club, they saw a critical need: students needed to be included in conversations about a technology that would fundamentally reshape their futures.

“AI was at the point where it could help people in every single major, in every single profession, in every single job,鈥� says Neary, a civil engineering major who co-founded United AI with Goodman, a biomedical engineering major, both in the (ECS). 鈥淲e realized this was no longer just a computer science thing.”

What started as a room of 10 people has grown into , a recognized student organization (RSO) with more than 100 members representing every single school and college and most majors. Since its fall semester launch, the club鈥檚 focus has been democratizing AI literacy and ensuring students from all disciplines have a seat at the table as this technology transforms society.

The group will host a on Saturday, April 25, from 1 to 5 p.m. in the K.G. Tan Auditorium in the National Veterans Resource Center at the Daniel and Gayle D鈥橝niello Building, featuring industry speakers, demonstrations and faculty research showcases.

Why Students Need Leadership in AI Development

For Goodman, the urgency became clear watching rapid AI development. “When I’m going through college, watching AI capabilities escalate, it can be disempowering鈥攁nd I figured my peers may be feeling the same way,” he says. “It felt threatening because there’s a small group of people making decisions about how the technology is being used, and others feel like they’re being left behind.”

That sense of being sidelined drove the co-founders to create what Neary describes as an empowerment space. “Something that we say a lot in the club is: don’t get used by AI, use AI to your benefit,” he says. “We’re the ones who are going into the workforce leading the charge and determining how we will use this technology now and into the future.”

The message resonated. Within weeks of tabling at campus events, students from ECS, the Maxwell School of Citizenship and Public Affairs, the College of Arts and Sciences, the Newhouse School of Public Communications, the Whitman School of Management and the College of Visual and Performing Arts were showing up to meetings, eager to understand how AI would affect their fields and futures.

Bringing Humanities and STEM Into Conversation

When Alex Kahn 鈥�27, a junior studying citizenship and civic engagement and political philosophy in the | , discovered United AI, he wasn鈥檛 looking for coding or technical skills, but was compelled by the policy implications of AI that were dominating news headlines. “AI was in every story, across every industry, and it felt like there was no escaping it and how it will affect you,” Kahn says.

As United AI’s recruitment director last fall, Kahn became instrumental in broadening the organization beyond its engineering roots. His approach focused on relevance rather than technical expertise. The interdisciplinary composition has transformed conversations within the club.

“Having people from different majors and disciplines means having that understanding that everyone’s mind works differently,” Kahn says. “The people who are writing code are not thinking the same way as the person majoring in fine arts, and having that creativity along with those technical skills, you’re able to build and think much differently.”

Goodman appreciates what non-engineering perspectives bring to the table. 鈥淎s conversations around AI progressed, I began asking, 鈥榃here are the artists? Where are the policymakers? Where are the humanities majors?鈥欌�� he says. “A lot of the population was not behind building this technology and still isn’t鈥攂ut how do we provide a space for them to learn and join the conversation?”

From Concept to Creation: Student Projects Take Shape

United AI goes beyond theoretical discussion to hands-on application. Through four-week project cycles, students receive funding, access to premium AI tools and mentorship to develop their ideas.

Paige Siciliano 鈥�29, a computer engineering major, led a second-place winning project during her first semester on campus. Her team’s AI-powered schedule builder, still under development, helps students manage their time by generating personalized daily plans based on individual learning styles, fixed commitments and flexible tasks.

For Siciliano and her teammates鈥擭eha Redda 鈥�29 and Ria Yagielski 鈥�29鈥攖he project provided more than AI experience. “It really helped us find a way into the community of 性视界, and it helped us feel like we belonged,” she says.

Building Community Around Shared Curiosity

Beyond projects and programs, United AI has cultivated what Kahn describes as “a school of thought on campus.” During a debate night last semester, members discussed everything from business applications to environmental impacts to personal usage philosophy, with some participants there simply to understand the technology rather than use it. “Being surrounded by club members and in this community of lifelong learners, we focus our educational efforts to not just learn the technical side, but also on practical application,” Kahn says.

Siciliano emphasizes the club’s welcoming atmosphere. 鈥淲e came in as first-semester freshmen, two weeks into school. It didn鈥檛 matter if we had no background knowledge in AI or all the knowledge in the world鈥攖hey create an atmosphere that makes you want to learn about it and continue to grow.”

To join United AI, . To learn more, follow the organization on or .

and the have partnered to launch an , a new self-paced credential designed to help business students evaluate sources, identify misinformation and apply research skills in a professional landscape increasingly shaped by artificial intelligence (AI).

The program, offered in collaboration with the Office of Microcredentials, is open to both Whitman undergraduate and graduate students and encourages the development of core skills in information literacy, which is a crucial competency for academic pursuits, and one that employers also describe as being essential. The skills learned also connect to the University’s of Information Literacy and Technological Agility and Critical and Creative Thinking.

“For Whitman students, the certificate fills a meaningful gap between classroom learning and professional readiness,鈥� says Assistant Director of Experiential Programs Roshawn Kershaw. 鈥淚t increases a student’s ability to find reliable information, assess its credibility and apply it with confidence. This is important for a business environment increasingly shaped by excess data and AI content. It sets them apart from others before they even realize. The certificate is now available to both undergraduates and graduate students, which means it can meet Whitman students wherever they are in their academic journey, reinforcing skills that will serve them from their first internship to the boardroom.”

To earn the certificate and digital badge, students take online self-paced tutorial modules that introduce them to key information literacy skills and library resources:

• Identifying Bias and Misinformation
• Types of Sources
• Evaluating Information
• Research as Process
• Search Basics, Part 1
• Search Basics, Part 2
• 性视界 Libraries Resources
• Student Guide to AI

鈥淚 am so excited to have these online tutorials become an official certificate and digital badge that is now available to both grads and undergrads,鈥� says Librarian for Business, Management and Entrepreneurship Steph McReynolds. 鈥淲e鈥檝e offered the tutorials as part of the program for years, and students have asked for a certificate to show employers their accomplishments in this area, and now we can provide that digital credential.鈥�

Information Literacy Librarian Kelly Delevan sees this certificate as an excellent template for the development of information literacy badges for other schools and colleges at 性视界. The certificate is even serving as a model beyond our institution, as a librarian from another university has recently reached out to use the certificate module categories at their own library.

Can an AI company take government money and still set limits on how its technology is used? That question is at the center of an ongoing dispute between the Pentagon and Anthropic, and 性视界 University professor Hamid Ekbia says it exposes fundamental tensions in how the AI industry operates.

Ekbia, founding director of the Academic Alliance for AI Policy, says the Pentagon鈥檚 demand that Anthropic either change its approach or forgo its lucrative contract is a vivid example of current federal policy. 鈥淲ith the bulk of public AI funding in the U.S. still coming from defense, companies either have to budge or shut themselves out from this unique source of money,鈥� Ekbia says.

While Anthropic has adjusted some safety policies, it has so far declined to allow its technology to be used for domestic surveillance or autonomous drones, a distinction Ekbia says matters.

“That is cause for celebration for any observer concerned about such applications,” he says. “But the question going forward is whether this will continue to be the case.”

Political and Economic Forces

Ekbia says the pressure on Anthropic reflects a broader shift in the federal government’s approach to AI regulation.

“The anti-regulatory policies of the Trump administration don’t leave much room for safety-oriented approaches to AI,” he says, adding that those policies push companies and oversight bodies toward “aggressive and often reckless behaviors in the name of innovation.”

Market competition makes the pressure worse. “The AI ecosystem is defined by furious competition among a few big players in a race to grab the lion’s share of the spoils in a rapidly growing industry,” Ekbia says. “The ‘moral economy’ of the AI industry is one of the jungle, where only the most reckless, ruthless, and aggressive behaviors are expected to be rewarded.”

Employees as a Wild Card

One factor that could shape the outcome is pressure from within Anthropic itself. Ekbia says employee resistance has played a meaningful role so far, with workers vocal during negotiations and leadership appearing to take that seriously.

But he cautions that employee influence is not guaranteed to last. “How critical will employees be in the future of the company given the current wave of white-collar聽under-employment, and how assertive will they be in expressing their resistance?” he says.

He outlines several other variables that will determine how the situation unfolds: whether competing AI companies are willing to fill the gap for the Pentagon, how hard the Trump administration continues to push for broad access to AI technology, and how well Anthropic can sustain itself financially without defense funding.

“The speed of change in these areas makes it hard to make solid predictions,” Ekbia says.

The Limits of Self-Regulation

Ekbia says the dispute ultimately tests a premise that Anthropic has staked its reputation on鈥攖hat a company can be both commercially successful and a responsible steward of powerful technology.

“In the absence of federal policy, Anthropic aspired to play that role in the industry,” he says. “What is happening shows the limited efficacy of that aspiration. Society cannot rely on the industry to self-police itself, despite even the best intentions.”

He connects that failure to a broader culture in Silicon Valley, where prominent figures publicly embrace “effective altruism”鈥攖he idea that profit and doing good can coexist.

“The case of Anthropic shows how much of an illusion this is,” Ekbia says. “As the old saying goes, you cannot have your cake and eat it too.”

Milton Santiago, assistant professor of visual communications in the S.I. Newhouse School of Public Communications, has received the 2026 Innovation in Teaching Award (BEA)鈥攐ne of the organization鈥檚 most prestigious honors for media educators.

A new minor in artificial intelligence science and engineering is designed to equip students with essential knowledge and skills in one of today鈥檚 most transformative fields. The minor, offered through the College of Engineering and Computer Science (ECS), will launch in the Fall 2026 semester.

New technologies such as Anthropic鈥檚 Claude and OpenAI鈥檚 ChatGPT are changing paradigms. The entire technology industry is pivoting toward the embrace of artificial intelligence. Coding agents are changing the way software is developed. Retrieval-augmented generation is changing the way companies manage data, and new systems promise further disruption. The new minor is designed to prepare students to thrive in this environment鈥攑roviding them with skills highly sought after by employers in the age of AI.

The 18-credit program combines core computing principles with specialized AI coursework, preparing graduates to navigate and contribute to the rapidly evolving landscape of artificial intelligence. It can be easily paired with other STEM majors.

The minor requires completion of 18 credits divided into two components:

Computing Foundations (nine credits): Students build essential technical skills through coursework focused on computational disciplines, establishing the groundwork necessary for advanced AI study and providing the programming and mathematical basis to understand advanced concepts such as language models and supervised machine learning.

AI Fundamentals and Programming (nine credits): These courses delve into artificial intelligence concepts, methodologies and applications, enabling students to develop expertise in this cutting-edge field. Courses include a strong focus on machine learning, using generative AI systems to create software and understanding large language models for various applications such as retrieval-augmented generation.

This minor is open to all University undergraduate students. It is designed for students seeking to enhance their primary degree with AI competencies.

Graduates of the program will possess key knowledge in artificial intelligence, positioning them competitively for careers in technology, research, data science and emerging AI-driven industries. As organizations across sectors increasingly integrate AI into their operations, this minor provides students with highly sought-after qualifications.

For more information about admission requirements and course offerings, students should contact their academic advisor or Priyantha Kumarawadu, associate teaching professor of electrical engineering and computer science and computer science undergraduate program director, at spkumara@syr.edu.

The growing day-to-day use of AI may come as no surprise, with its integration into daily tools like email, smart devices and social media.

But it鈥檚 also becoming more common for people to engage with AI for emotional or romantic companionship. Newhouse School of Public Communications associate professor 鈥攚ho studies youth, media and sexual health鈥攄ecided to examine the trend. Hearing from young people鈥攁nd seeing 鈥攁bout how they use AI to build relationships or understand their own human connections prompted her research.

鈥淚f I’m going to continue to research how media play a role in young people’s lives, particularly as it relates to their sexual health and romantic relationships, AI chatbots and companions are going to have to be part of that conversation,鈥� she says.

Ortiz and her colleagues surveyed young people to learn how they鈥檙e using AI chatbots or companions for romantic, emotional or sexual purposes and how that use relates to their own romantic boundaries and communication.

鈥淥ne of my questions was, 鈥楬ow might use of AI for romantic companionship result in helpful or harmful outcomes?鈥欌�欌�� Ortiz says. 鈥淔or example, could practicing communication with an AI companion help someone communicate with a human partner, such as practicing how to flirt or how to say things they might feel uncomfortable saying?鈥�

The Survey

Ortiz and her colleagues surveyed 1,500 18-to 21-year-olds. They ended up with about 360 respondents reporting they used AI for romantic companionship.

性视界 two-thirds of the 360 said they used AI companions similar to a long-term romantic relationship, communicating over a period of time rather than a single, one-off interaction.

Ortiz says quite a few reported using AI to 鈥減ractice鈥� how to engage in their human relationships. One participant shared they were having some problems with their romantic partner, and they used AI to roleplay how they might cheer up their loved one or help them feel better.

鈥淭his respondent said it gave them some guidance for what to do,鈥� Ortiz says. 鈥淭hen there were some people who said, 鈥業t helped me figure out how to flirt. It helped me work through some of the awkwardness of communication.鈥� So at least some young people are using these companions to practice or get a sense of what it would be like when they take it to a human relationship.鈥�

Areas of Concern聽

Ortiz says one concern she and her colleagues observed was that some AI companions default to sexually aggressive language or exchanges that do not follow a constructive, consenting back-and-forth.

鈥淭his is concerning regardless of what age you are, but we are particularly concerned for young people who are still learning how to communicate about consent and boundaries,鈥� Ortiz says.

What they found is some of the apps, if given an indication the user was interested in sexual or romantic communication, would almost immediately become sexually aggressive.

Ortiz says those responses are a red flag that AI companions can model unhealthy, abusive communication, an important element to further examine and include in discussions about AI companions.

Experiencing Stigma聽

The survey asked participants about their emotional connection with AI companions and whether they felt the tool understood their emotions, among other questions probing the relationship between the young adult and the technology.

What Ortiz says she found is some of the young people did express a strong connection with the AI companion, with some listing loneliness as a motivation for usage.

Even with the belief that AI could be a 鈥渟afe space,鈥� Ortiz says her survey indicates there is still stigma around using AI tools for romantic or sexual purposes.

鈥淢any in the survey reported that they thought using AI for romantic companionship was weird, unacceptable, not a normal thing to do,鈥� Ortiz says. 鈥淢ost of the respondents didn’t think this was a common behavior among people their age, but you can see there is a good chunk of young people who are using it for these purposes.鈥�

What Should Be Asked Next

Ortiz says there is not a clear indication that using AI companions or chatbots for romantic companionship is leading to healthier outcomes for most users.

鈥淯nfortunately, the results show that, for some users, engaging with these AI companions has the potential to be related, not necessarily causing, but related to less healthy romantic beliefs and behaviors,鈥� she says.

Ortiz hopes her work can serve as a warning sign to people creating companion apps or platforms like ChatGPT that boundaries and guardrails should be built-in so users can engage in healthy, safe ways.

People are building real relationships with AI companions, and the goal should be to understand and ensure there are healthy outcomes, without being too judgmental, she says.

鈥淎I companionship is not going away,鈥� Ortiz says. 鈥淪o the question should be, how can this be used in more helpful than harmful ways if we know young people are going to use it? It’s just another tool for young people to help make sense of themselves, and we should be open to understanding that if we want to help them build healthy relationships.鈥�

When doctoral student Jamie Gentry G鈥�27 covered politics as a local news reporter for the weekly Navarre Press in northwest Florida, she turned potentially complicated issues into easy-to-understand stories.

But Gentry was amazed at how often people would misinterpret, misconstrue or misremember the information presented in her articles. She overheard many conversations in person and online where citizens, equipped with this misinformation, carried out emotional arguments on a topic using incorrect information.

鈥淚 started wondering why I wasn鈥檛 able to reach as many people as I could with the actual facts of a story,鈥� Gentry says. 鈥淚t was frustrating because my job is to give people the best possible information. People need good information to make good decisions, and journalists are supposed to do that. But I found the system wasn鈥檛 working.鈥�

Gentry knew there was a disconnect between how political news was being reported and how it was being talked about in her community. She vowed to become part of the solution.

How to Fix a Broken System

Driven by her reporting experiences, Gentry transitioned from journalism to higher education and began pursuing a doctoral degree in mass communications from the .

With a grant from the University’s , Gentry鈥檚 ongoing research explores how artificial intelligence (AI) tools used by journalists impact how politics are discussed online and in the real world.

Gentry is comparing how people respond to and discuss a complicated news topic among their communities and on their social media channels under two different scenarios.

Out of 400 online survey respondents, one group is tasked with reading a traditional news story about unemployment, while another digests the information with the help of an AI-generated key takeaways breakout box. Half of the participants are told to share their impressions of the article with someone they know face-to-face, while the other half are tasked with sharing a post about the topic on social media.

At each step, from the journalist sharing their reporting to the survey participant consuming the content to the person receiving the news, there鈥檚 an opportunity for the message to change from the original reporting.

鈥淕enerally, people tend to accept facts, but we still see arguments over facts online, and we see that people become very polarized,鈥� Gentry says.

An important trend in the political communications research field鈥攃ombining the study of media and political science鈥攊s examining how, in an increasingly polarized country, being divided politically impacts the quality of political reporting.

Especially during this 鈥渆xplosion of media choice鈥� where people have more ways to consume the news, Gentry says this increase in choice means people are opting for stories they want to consume that align with their political ideology.

鈥淭hat has a real impact on how people engage with politics and how they interpret the news they receive,鈥� Gentry says when identifying an area for future research. 鈥淚t鈥檚 not so much that people are blatantly believing misinformation and don鈥檛 care about facts. It鈥檚 more that partisanship is impacting how people receive messages and what stories they do and do not see.鈥�

Can AI Be Trusted?

As informers, journalists are charged with breaking down complex topics into digestible content, and they make decisions about what information to include, which sources to interview and which stories to cover.

When she was covering the news, Gentry says it was easy to think she knew what the most important angles were, but as more journalists use AI to produce story summaries, Gentry says it鈥檚 natural to wonder whether AI can convey this important information.

鈥淛ournalists influence how people learn about and understand a subject matter. Should we be trusting these AI tools to reliably make decisions about what is the most important part of a story?鈥� Gentry says. 鈥淲hatever AI decides is the most important snippet of information is being pushed out and that has real implications for how people are getting the news and what they actually know about a story.鈥�

Gentry expects to receive data from her survey participants later this semester. Among her anticipated findings: story summaries make the facts more accessible and easier to process, retain and share.

鈥淢y goal is to make journalists better by giving them the tools to better understand how their work impacts the public,鈥� Gentry says. 鈥淏y sharing data on what works and what doesn鈥檛, hopefully we can make big improvements in the way the news is shared.鈥�

How are law firms currently applying Artificial Intelligence in the workplace to maximize client services? What are the ethical implications of using AI in the legal field? How will AI impact the current role of lawyers, and what new jobs may emerge? Should AI be regulated, and, if so, how?

These were just some of the questions addressed during AI and the Virtue of Law, a one-week in-person residency held at the in August designed for students in the , with participation also open to on-campus students. This deep-dive into AI was created and facilitated by聽.

鈥淚 think AI will significantly transform law school education and the practice of law,鈥� says Graves, noting that he sees AI as a means of more effective information sharing but also recognizes that many are 鈥渢errified鈥� thinking that this technology could replace them.

鈥淲e have to think about being nimble now because the essential human role today will likely be an AI role in just a few years, and we don鈥檛 want to be left behind. Through this residency, I wanted to help demystify generative AI because, used properly, it can be an extraordinary tool,” Graves says.

Graves, who has taught in the JDi program for the past five years, has a unique blend of expertise in design, development and delivery of accessible and legal education in an online learning environment and 21^st century, technology-leveraged law practice.

A graduate of the University of Colorado Law School, Graves taught the technology-leveraged delivery of legal services at the Touri Law Center for 14 years. Before that, he worked in private practice with Chrisman, Bynum & Johnson PC in Colorado, and as a judicial law clerk for the U.S. Court of Appeals.

First Time Residency a Popular Draw

Logan Gorg L鈥�26 is a JDi student living in Pennsylvania who made the trip to campus to attend the AI residency. She has worked as a paralegal at the law firm of Ross & Ross LLC for the past 10 years and is looking to focus on real estate and probate law upon graduation.

鈥淚 learned so much about what AI is, and the information at the residency helped to dispel some of the fears and focus more on where the profession is going,鈥� Gorg says. 鈥淪itting in a room with a group with diverse backgrounds and experiences talking about whether AI was doom or salvation was so interesting. I think the residency showed us that AI is unavoidable, but, if we get out in front of it, we can reap some of the benefits in the legal profession.鈥�

Graves had been contemplating developing a semester-long course in AI for the JDi program, but ultimately he decided that the lightning speed of the technology would be better suited for a short-form, concentrated residency where students with different levels of familiarity could join together to think about being nimble and adapting to technology that is already changing the way the legal field operates.

AI Voice-Driven Technology Used to Teach, Demonstrate Abilities

Coincidentally, the residency took place just as ChatGPT launched Advanced Voice Mode, a significant upgrade that allows for natural, real time conversations using AI. Graves used 鈥淢ax,鈥� as he named the voice-driven AI technology, to help co-teach the residency and answer students鈥� questions directly.

鈥淲e would have a discussion, and I would say, 鈥楳ax, what do you think?’” Graves says. 鈥淎t first students were uncomfortable with it, but once Max started responding and asking them questions using the Socratic method, they started to see how fascinating a learning AI tool could be.鈥�

Approach to AI in the Law Resonates With JDi Students

Jenny Cameron L鈥�27, who co-owns VIP Marinas with her husband in Florida, decided to enroll in law school to bring a legal perspective to her family business. She, too, attended the AI residency and walked away amazed.

鈥淗onestly, it was one of those residencies that was life changing,鈥� Cameron says. 鈥淏efore I attended, I was on the fringes of AI, barely using ChatGPT, but since then I鈥檝e been using AI extensively in some form. Part of law school is practicing and knowing how to use AI better and faster, and what I learned at the residency was eye opening. I commend the College of Law and Professor Graves for taking the lead on this and helping guide us on how we should be approaching this technology.鈥�

Another participant was Bryan Beene 鈥�26 a high school government teacher from Texas, who is pursuing law school to prepare for a second career once he retires. He hopes to work as a lawyer in the education or church law space.

鈥淚 registered for this AI residency for two reasons: one because Professor Graves was teaching it, and he is one of the best professors I鈥檝e ever had; and two, because I had never used AI except for Google searches, and I knew a lack of knowledge around this technology would be a detriment in representing a client,鈥� Beene says.

Beene noted that he enjoyed learning more about the use of the available tools, as well as discussing the legal and ethical issues, and how regulations and the law are often not keeping up with this fast moving technology.

The newly introduced AI and the Virtue of Law residency received 鈥渋ncredible feedback鈥� from students, says Graves, who believes this is a topic that should be revisited once a year.

鈥淭his is not a static course, as the technology is changing continuously, but I think the approach resonated well with the students, not only by teaching them skills but by allaying some of their fears while also emphasizing to them that AI technology in the legal field is advancing fast and furiously. So they need to prepare now,” Graves says.

If you have a kid or teen at home, you’ve probably heard of Roblox鈥攁n online gaming platform where millions of users create, share and play games in a virtual world with its own currency.

The company’s new age verification requirement for users under the age 13 sounds like a smart safety measure, but it actually highlights a critical vulnerability affecting gaming platforms, social media sites and any technology company relying on camera-based authentication to protect minors online.

, professor of electrical engineering and computer science at 性视界 University and a Fellow of the Institute of Electrical and Electronics Engineers (IEEE) and American Association for the Advancement of Science, warns that these systems may create a false sense of security. His research focuses on cybersecurity, machine learning and biometrics鈥攊ncluding touch-based and network-based authentication systems.

While AI-powered facial recognition can estimate age within two to three years by analyzing wrinkles and skin condition, Phoha warns the systems are “highly susceptible to spoofing.” Simple presentation attacks鈥攍ike a printed photo, a smooth mask to mimic younger skin or even a silicone dummy face鈥攃an fool verification systems, especially when users authenticate through their own laptop or phone cameras lacking sophisticated liveness detection.

As more platforms implement age verification requirements to comply with child safety regulations, Phoha says the technology may create a false sense of security.

Phoha answers three questions about the effectiveness of this technology and is available for interviews on biometric authentication, age verification technology and cybersecurity topics.

What are the the challenges and effectiveness of systems that collect biometric information via palms, eyes, face, etc.?

“Typically, age estimation from a face using a camera relies on features such as wrinkles, skin condition and sagging. The deep learning (AI) models are trained on millions of labeled faces to recognize age-specific features on a face. The training labels adjust for gender, race, etc. Although literature supports high accuracy in age estimation, ranging from a difference of two to three years, I believe the methods are highly susceptible to spoofing.

“Also, the accuracy of authentication methods depends on lighting conditions, pose etc. Typically, the features of the account for the variations in light, pose, etc. Literature also suggests using eyes and typing patterns, including the use of language to predict age, although these studies are not common.”

Tell us more about authentication methods and their reliability for protecting minors online.

“Most authentication is done through deep learning and pattern recognition algorithms. However, all camera-based methods are susceptible to spoofing. Earlier experiments of presentation attacks showed that a copy of a face printed on paper was able to fool the system, because the proposed system by Roblox will use 3D, which can also be fooled, for example, by a simple skin colored (smooth mask to appear as younger, healthier skin). Or a dummy 3D face made of silicone. Because the verification of an individual鈥檚 face may be done using the individual鈥檚 (laptop or desktop) camera, it will likely not have sophisticated hardware to test for liveness, such as twitching, blood flow, sweat glands, etc.”

What are the broader implications for how tech platforms balance safety with privacy concerns?

“If the features of the face (and not the images of the face) are stored (or destroyed immediately) after collection, one can hope that no personally identifiable information will be discernible.”