Hart Flags Artificial Intelligence Media to Vet

Hart urges verification of realistic AI media

WCTI News 12 published a report on October 16, 2026, in which ECU Assistant Professor of Computer Science Dr. David Hart said artificial intelligence tools could become more personalized, helping people organize their days through simple conversations or requests.

WCTI News 12 said artificial intelligence is becoming part of everyday technology, from search engines and apps to chatbots. Researchers and policymakers are weighing its potential benefits against concerns about misinformation, safety and how much authority the technology should have, the station said.

Hart said people should be cautious about realistic AI-generated images, voices, text and video, WCTI News 12 reported. "The key will be, sadly, almost to trust nothing you see until you can verify it with some kind of external source," Hart said. He encouraged people to look for reliable confirmation, especially when content seems extraordinary, the station reported.

Hart also said AI does not have its own desire to take over the world. Instead, he said, it tries to optimize the task people give it, which makes human choices and oversight central to how the technology is used. "The key will be making sure that we're using AI for the good as many ways as other people are trying to use it for bad," Hart said.

Anthropic researchers voice severe safety concerns

Evan Hubinger, an alignment science lead at Anthropic, said he personally believes there is more than a 10% chance that AI could kill all humans within the next decade, WCTI News 12 reported. The station said the figure is Hubinger's personal estimate, not a certain outcome.

Researchers do not yet have a plan to solve the challenge of aligning superintelligent AI with human goals, Hubinger said, according to the station.

Jacob Coxon, a former Anthropic researcher, said companies are moving too fast, WCTI News 12 reported. Coxon criticized the industry's push toward AI systems that could improve themselves.

Anthropic was among the companies that signed a voluntary accord pledging internal safety controls, outside reviews and board-level oversight, the station said. It added that AI is also being used in military analysis.

IMF chief calls for AI guardrails in Bangkok

IMF Managing Director Kristalina Georgieva said Friday that guardrails are urgently needed to contain potential risks from the boom in artificial intelligence, according to an AP report that Greenwich Time published on October 16, 2026, from the IMF and World Bank annual meetings in Bangkok.

Georgieva likened the technology to a Formula 1 race car that might go off the track and run people over, Greenwich Time reported. "We are racing, and the question is how do we put barriers? So if a car was to fly out it doesn't kill people," Georgieva said. She added that the industry's work is important, but that "if the car runs over people, that gift may not be so worthwhile."

Georgieva spoke on a panel on the impact of AI on the world economy, alongside experts including Anthropic co-founder and head of policy Jack Clark, Greenwich Time reported. The AP story described Clark as among technology industry leaders pushing for faster action on regulating AI.

The Bozeman Daily Chronicle ran the AP story on October 16, 2026. Clark noted that toys are more regulated than AI and stressed the need for policy frameworks to control AI development, the newspaper reported. An IMF study found AI could transform or create 10% of jobs, but repetitive tasks may disappear, the newspaper said. Georgieva also highlighted the need for access to electricity and digital networks to maximize the benefits of AI.

California campuses weigh AI and education

University Times reported on October 16, 2026, that questions about AI's benefits, ethical implications and government oversight are gaining attention at California universities.

Newsom said in September 2026, in a Los Angeles Times story the University Times cited, "We regulate your barber or your sandwich shop more than we regulate [an AI company] telling us that they may be ending humanity." The University Times also described a call for stronger oversight of advanced AI systems, including a potential emergency shutoff mechanism, commonly referred to as a "kill switch," that could disable certain AI models if they pose serious safety risks.

An instructor identified in the paper as Sun said many of his concerns involve academic integrity while he teaches students in his department how to use AI for their fields of work. "AI just kind of sprung up on us," Sun said. "I think this is doing great harm in teaching and learning," he said.

Verdugo, an assistant professor of computer science who said he leads the university's Multidisciplinary Intelligence for Novel Data Solutions Lab, took a different view, the University Times reported. "Our Curriculum should integrate AI use across subjects showing how it can boost productivity and augment human capabilities rather than replace them," Verdugo said.

Meer commentary lists everyday AI uses

Meer published a commentary on October 15, 2026, saying AI sorts emails, powers face recognition and assists with navigation on phones. Meer wrote that AI handles data analysis and repetitive tasks, which lets workers save time.

Meer wrote that AI today is narrow, specialised and data-driven, excelling in predefined tasks but lacking the broad, adaptable intelligence of humans.

The commentary also said relying too much on large language models can weaken human problem-solving and writing skills. Meer wrote that leaders need to embrace the technology, learn to harness its potential and develop use cases for their businesses.

Tron's take

My take is that the day's news gives owners things to weigh, not things to buy. Hart's verification advice is the item with the most direct everyday use. That is my reading of the news, not a reported result.

I would turn Hart's advice into a written rule: any voicemail, video call, invoice or vendor email that asks to move money, change bank details or reset credentials gets confirmed through a phone number already on file. Convincing-looking media is the case Hart described, and the check costs one phone call.

I would also keep adoption deliberate. Meer's examples, email sorting and data analysis, are established uses, and they do not depend on this week's headlines. Owners can follow the news to time decisions without acting on every item. For related coverage of the rules debate, see XL.net's earlier report, Microsoft's Brad Smith Urges AI Rules to Weigh.