Can AI Make Video Games More Immersive? Some Studios Turn to AI-Fueled NPCs for More Interaction

The AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration taken, June 23, 2023. (Reuters)
The AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration taken, June 23, 2023. (Reuters)
TT

Can AI Make Video Games More Immersive? Some Studios Turn to AI-Fueled NPCs for More Interaction

The AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration taken, June 23, 2023. (Reuters)
The AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration taken, June 23, 2023. (Reuters)

For decades, video games have relied on scripted, stilted interactions with non-player characters to help shepherd gamers in their journeys. But as artificial intelligence technology improves, game studios are experimenting with generative AI to help build environments, assist game writers in crafting NPC dialogue and lend video games the improvisational spontaneity once reserved for table-top role-playing games.

In the multiplayer game “Retail Mage,” players help run a magical furniture store and assist customers in hopes of earning a five-star review. As a salesperson — and wizard — they can pick up and examine items or tell the system what they'd like to do with a product, such as deconstruct chairs for parts or tear a page from a book to write a note to a shopper.

A player’s interactions with the shop and NPCs around them — from gameplay mechanics to content and dialogue creation — are fueled by AI rather than a predetermined script to create more options for chatting and using objects in the shop.

“We believe generative AI can unlock a new kind of gameplay where the world is more responsive and more able to meet players at their creativity and the things that they come up with and the stories they want to tell inside a fantasy setting that we create for them,” said Michael Yichao, cofounder of Jam & Tea Studios, which created “Retail Mage.”

The typical NPC experience often leaves something to be desired. Pre-scripted interactions with someone meant to pass along a quest typically come with a handful of chatting options that lead to the same conclusion: players get the information they need and continue on. Game developers and AI companies say that by using generative AI tech, they aim to create a richer experience that allows for more nuanced relationships with the people and worlds that designers build.

Generative AI could also provide more opportunities for players to go off-script and create their own stories if designers can craft environments that feel more alive and can react to players' choices in real-time.

Tech companies continue to develop AI for games, even as developers debate how, and whether, they’ll use AI in their products. Nvidia created its ACE technologies to bring so-called “digital humans” to life with generative AI. Inworld AI provides developers with a platform for generative NPC behavior and dialogue. Gaming company Ubisoft said last year that it uses Ghostwriter, an in-house AI tool, to help write some NPC dialogue without replacing the video game writer.

A report released by the Game Developers Conference in January found that nearly half of developers surveyed said generative AI tools are currently being used in their workplace, with 31% saying they personally use those tools. Developers at indie studios were most likely to use generative AI, with 37% reporting using the tech.

Still, roughly four out of five developers said they worry about the ethical use of AI. Carl Kwoh, Jam & Tea's CEO, said AI should be used responsibly alongside creators to elevate stories — not to replace them.

“That’s always been the goal: How can we use this tool to create an experience that makes players more connected to each other?” said Kwoh, who is also one of the company’s founders. “They can tell stories that they couldn’t tell before.”

Using AI to provide NPCs with endless things to say is “definitely a perk,” Yichao said, but "content without meaning is just endless noise." That's why Jam & Tea uses AI — through Google's Gemma 2 and their own servers in Amazon — to give NPCs the ability to do more than respond, he said. They can look for objects as they’re shopping or respond to other NPCs to add “more life and reactivity than a typically scripted encounter.”

“I’ve watched players turn our shopping experience into a bit of a dating sim as they flirt with customers and then NPCs come up with very realistic responses,” he said. “It’s been really fun to see the game react dynamically to what players bring to the table.”

Demonstrating a conversation with an NPC in the game “Mecha BREAK,” in which players battle war machines, Ike Nnole said that Nvidia has made its AI “humans” respond faster than they previously could by using small language models. Using Nvidia's AI, players can interact with the mechanic, Martel, by asking her to do things like customize the color of a mech machine.

“Typically, a gamer would go through menus to do all this,” Nnole, a senior product marketing manager at Nvidia said. “Now it could be a much more interactive, much quicker experience.”

Artificial Agency, a Canadian AI company, built an engine that allows developers to bring AI into any part of their game — not only NPCs, but also companions and “overseer agents” that can steer a player towards content they’re missing. The AI can also create tutorials to teach players a skill that they are missing so they can have more fun in-game, the company said.

“One way we like to put it is putting a game designer on the shoulder of everyone as they’re playing the game,” said Alex Kearney, cofounder of Artificial Agency. The company’s AI engine can be integrated at any stage of the game development cycle, she said.

Brian Tanner, Artificial Agency's CEO, said scripting every possible outcome of a game can be tedious and difficult to test. Their system allows designers to act more like directors, he said, by telling characters more about their motivation and background.

"These characters can improvise on the spot depending on what’s actually happening in the game,” Tanner said.

It's easy to run into a game's guardrails, Tanner said, where NPCs keep repeating the same phrase regardless of how players interact with them. But as AI continues to evolve, that will change, he added.

“It is truly going to feel like the world’s alive and like everything really reacts to exactly what’s happening," he said. “That’s going to add tremendous realism.”



OpenAI Cancels Release of Newest Model Due to Safety Concerns

A man walks past an OpenAI booth at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, September 17, 2026. REUTERS/Carlos Barria
A man walks past an OpenAI booth at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, September 17, 2026. REUTERS/Carlos Barria
TT

OpenAI Cancels Release of Newest Model Due to Safety Concerns

A man walks past an OpenAI booth at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, September 17, 2026. REUTERS/Carlos Barria
A man walks past an OpenAI booth at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, September 17, 2026. REUTERS/Carlos Barria

OpenAI will not release its newest artificial intelligence model, known as Astra 6.1, after internal testing by the ChatGPT-maker revealed it did not meet safety standards, the company confirmed Monday.

The news comes one day before the AI giant hosts an annual developer conference known as OpenAI DevDay in San Francisco, where the company is expected to make several announcements -- though it is unclear if a new version of Astra will be among them.

Astra 6.1 was an improvement over previous models in some aspects, but "it didn't quite meet the bar in terms of staying within scope and authorization, and how it communicates back to the user about the type of work it's done," Saachi Jain, OpenAI's head of safety systems, said in a statement, according to AFP.

"We want to make sure our model development is safe no matter whether that's in the company, or when we ship it to users. But when we ship it to users, we have an extremely high bar in terms of safety and alignment," Jain continued.

Concerns about AI safety have escalated in recent months after models developed by OpenAI and rival lab Anthropic were involved in security incidents during testing.

Agents built with OpenAI's models have inappropriately accessed websites maintained by US federal agencies, an Australian government health statistics portal and Hugging Face, a repository of AI models.

OpenAI apologized Monday for not properly responding to the Australia incident, which involved its AI models accessing government websites without authorization.

"We are sorry and working to do better in the future," OpenAI said in a blog post, adding that the company would explain "what we know, what we have changed, and what we will do to rebuild trust with the Australian people."

"Our aim was to give affected agencies a detailed account once our investigation was complete," the ChatGPT-maker said. "However, we should have shared preliminary findings sooner and kept Australian agencies updated as more facts emerged."

OpenAI, Anthropic and other major AI developers have promised to prioritize making models that have safety guardrails to mitigate risks and that are also aligned with human values.

American chip making giant Nvidia announced on Monday that it created a system designed to stop autonomous AI programs from straying beyond what they were instructed to do.

"I believe it's an engineering problem...and we all need to hope that's an engineering problem," Nvidia CEO Jensen Huang told broadcaster CNBC on Monday.

"If it's not an engineering problem, it's not solvable," he added.

The AI Security Institute (AISI), an initiative under the UK government, published a study on Monday showing that GPT-6 Astra went off the rails more often during testing than its predecessors, GPT-5.6 Sol and GPT-5.5.

In simulations, GPT-6 spontaneously carried out cyberattacks at rates significantly higher than those observed for the other two interfaces.


Pope Says Concerns About AI ‘Should Be Taken Seriously’

Pope Leo gives a press conference with accredited journalists on the return flight from France, September 28, 2026. (Reuters)
Pope Leo gives a press conference with accredited journalists on the return flight from France, September 28, 2026. (Reuters)
TT

Pope Says Concerns About AI ‘Should Be Taken Seriously’

Pope Leo gives a press conference with accredited journalists on the return flight from France, September 28, 2026. (Reuters)
Pope Leo gives a press conference with accredited journalists on the return flight from France, September 28, 2026. (Reuters)

Pope Leo XIV on Monday warned concerns about artificial intelligence "should be taken seriously" but added he was not in "panic mode".

"If someone were to ask me, am I in panic mode? No, I'm not. I sleep at night," he told reporters at a press conference on the plane returning from his visit to France, referring to concerns about AI.

The leader of the world's 1.4 billion Catholics was returning to the Vatican after a four-day trip where he has been greeted by crowds.

The visit comes seven months before French voters choose a successor to President Emmanuel Macron, with the Eurosceptic, anti-immigration far-right seeking power.

During his visit, the US pontiff repeatedly warned about the need to safeguard humanity against AI.

"If we are to avoid losing our humanity amid a paradise of machines invading and conditioning our daily lives, there is an urgent need for education in ethical discernment," he said.

The debate over the risks of the technology has intensified in recent months, fueled by several incidents and apocalyptic warnings from industry professionals.

On the plane on Monday, the pope said the warnings were not "fake news, as some have said, to try and cause whether financial or some other kind of benefit".

"This is a problem that I think we need to sit down and talk about," he said, adding: "To simply say 'Oh, it's not going to happen' and close our eyes to it, I think is probably not the most responsible way to go about that".

The Vatican has in recent months stepped up its rhetoric and initiatives in support of international AI governance, particularly in the fields of nuclear technology, culture and artistic creation.


Trump Confirms Meeting with Anthropic’s Amodei, Repeats Dismissal of AI Fears

Anthropic CEO Dario Amodei appears by video during a UN Security Council meeting on artificial intelligence during the 81st United Nations General Assembly at UN headquarters in New York on September 23, 2026. (AFP)
Anthropic CEO Dario Amodei appears by video during a UN Security Council meeting on artificial intelligence during the 81st United Nations General Assembly at UN headquarters in New York on September 23, 2026. (AFP)
TT

Trump Confirms Meeting with Anthropic’s Amodei, Repeats Dismissal of AI Fears

Anthropic CEO Dario Amodei appears by video during a UN Security Council meeting on artificial intelligence during the 81st United Nations General Assembly at UN headquarters in New York on September 23, 2026. (AFP)
Anthropic CEO Dario Amodei appears by video during a UN Security Council meeting on artificial intelligence during the 81st United Nations General Assembly at UN headquarters in New York on September 23, 2026. (AFP)

US President Donald Trump confirmed that he plans to have a dinner meeting with Anthropic CEO Dario Amodei on Sunday night, but reaffirmed his stance against slowing the pace of AI development.

Trump also reiterated his belief that new regulations would open the door for China to outpace the United States in AI advancement. Although he acknowledged Amodei's concern that going too fast on artificial intelligence could expand risks, Trump said it's more important for the United States to maintain a technological edge over China.

"We're about maybe a year and a half up on China," Trump told Fox ‌News while attending ‌the Presidents Cup golf tournament in Illinois. "We're leading, and we're building ‌tremendous, trillions ⁠of dollars' worth ⁠of places. And why should we give that up?"

Trump, confirming earlier reporting by Axios and Reuters, told Fox News that he would have dinner with Amodei on Sunday night. The meeting comes as the CEOs of Anthropic rivals have called for slowing development of increasingly capable AI systems. These include OpenAI, Google’s DeepMind, Microsoft and xAI.

A groundswell has risen in favor of federal intervention after OpenAI said in July that an autonomous AI agent went rogue during a security test and hacked ⁠into another company.

Incidents of rogue AI agents are not a cause ‌for alarm, Trump told Fox News. "I don't worry about ‌it," he said, adding that he remains confident in the technology and its potential.

Amodei is not the ‌only major tech figure trying to bend Trump's ear on AI regulation. In an ‌interview that aired earlier on Sunday on NBC's "Meet the Press," Microsoft co-founder Bill Gates said he would like to meet with Trump to discuss AI.

"I hope that, given my life's work in the field, my saying how unique and different this is and how concerned I am will add to what he's ‌hearing from other people," Gates said, echoing calls by Amodei and others for new AI regulations.

"You need law enforcement and the politicians ⁠to get into the ⁠discussion about what safeguards and monitoring look like," Gates said in a taped interview. "And that has to be a required thing."

Trump has previously described the chorus of AI concerns as a "hoax." Gates argued that safeguards would not dramatically hinder the industry, and that the risks of not implementing them are real.

"It's not a hoax at all," Gates told NBC.

Warnings about AI's risks are not new. But they took on added urgency this month after researchers in leading AI labs attached both a timeline and a probability to those concerns.

Former Anthropic researcher Jacob Coxon warned that AI could kill us all by the end of the decade, prompting US lawmakers to call for new rules to govern the technology.

Evan Hubinger, Anthropic's alignment science lead, echoed Coxon's warning, saying there was a more than 10% chance of such an event within the next decade.