A large majority of Americans believe artificial intelligence companies need to do more to prevent potentially serious harm from increasingly powerful AI systems, according to a new national survey.

The poll found that 73% of respondents were concerned that AI companies had not implemented enough safeguards to protect society from serious risks. The findings also indicate declining public confidence in the technology’s impact and limited support for allowing AI companies to regulate themselves without stronger government involvement.

Public Concern About AI Is Rising

The four-day survey found that 73% of participants believed AI companies had not gone far enough in establishing protections against serious societal harm.

At the same time, 39% said artificial intelligence was having a negative impact on society, up from 36% in the previous month. That represented the highest level recorded in the survey series.

Only 11% viewed AI’s impact as positive, while the remaining respondents were either uncertain or did not provide an answer.

The results reflect growing public debate as AI systems become more capable and increasingly integrated into workplaces, businesses and everyday services.

Americans Show Limited Confidence in Industry Self-Regulation

The survey also indicated that many Americans want a stronger government role in establishing AI safety requirements.

Most respondents supported a significant role for federal authorities in setting standards for AI safety, reflecting concerns about whether technology companies can effectively oversee their own development and risk-management practices.

The issue has become a major part of the broader discussion surrounding artificial intelligence, particularly as companies deploy increasingly capable models and autonomous systems.

Majority Support Slowing AI Development

The poll found that 55% of respondents considered slowing AI development to be a good thing.

By comparison, 13% said slowing development would be negative, while the remaining participants were unsure or did not answer.

The survey also found that approximately 69% had followed reports about major AI companies calling for a slowdown in development amid concerns about safety and oversight.

The findings highlight an ongoing debate over whether AI development is advancing faster than the safeguards designed to manage its potential risks.

Cybersecurity and Privacy Remain Major Concerns

Public concern about AI is not limited to long-term questions about the technology. Cybersecurity, privacy and potential misuse are also contributing to the debate.

The survey report pointed to incidents involving increasingly autonomous AI agents as an example of the challenges facing developers.

One concern is that AI systems capable of interacting with external tools, networks and services could potentially behave in unexpected ways or be misused by attackers.

At the same time, supporters of AI point to potential benefits including increased productivity, scientific research, healthcare applications and improvements across financial and industrial sectors.

AI Researchers Also Raise Safety Concerns

Warnings from people working within the AI research community have added to the public discussion.

Some researchers have expressed concerns about the consequences of increasingly capable AI systems and questioned whether existing safety measures will be sufficient as models become more autonomous and powerful.

These concerns have contributed to a broader debate about the pace of AI development and the responsibility of companies building advanced systems.

However, the potential economic and scientific benefits of AI continue to drive significant investment and research, making the balance between innovation and safety an ongoing issue.

Safety Takes Priority Over Global AI Competition for Many Americans

The survey also examined whether Americans place greater importance on developing AI safely or maintaining U.S. leadership in global AI competition.

According to the findings, 73% said safe and responsible AI development was more important, while 23% placed greater importance on ensuring the United States remains ahead of other countries in AI development.

The result suggests that, among those surveyed, concerns about safety outweighed concerns about maintaining an international technological advantage.

Survey Details

The poll was conducted online among 1,277 U.S. adults.

It reported a margin of error of approximately three percentage points in either direction.

Because the survey represents a snapshot of public opinion at a particular point in time, attitudes toward AI may change as new technologies, regulations and high-profile incidents emerge.

Public Debate Over AI Is Evolving

The survey points to a growing tension between rapid advances in artificial intelligence and public expectations for stronger safeguards.

AI systems are increasingly being used for work, research, communication and decision-making, while concerns about cybersecurity, privacy, misuse and autonomous behavior continue to develop.

The findings suggest that many Americans want technological progress to continue while placing greater emphasis on safety and accountability.

As AI capabilities expand, the debate is likely to focus increasingly on whether regulatory frameworks and industry safeguards can keep pace with the technology.

Three Indian cybersecurity researchers used Anthropic’s Claude AI models to uncover and demonstrate security weaknesses affecting systems connected to OpenAI, highlighting how advanced artificial intelligence can significantly accelerate vulnerability research.

The researchers—Harsh Jaiswal, Mohan Pedhapati and Rahul Maini of cybersecurity startup Hacktron AI—reported their findings to OpenAI after identifying vulnerabilities that potentially exposed user accounts, connected services and parts of the company’s internal infrastructure.

OpenAI subsequently awarded the researchers a $6,500 bug bounty, equivalent to roughly ₹6.22 lakh.

Who Discovered the OpenAI Security Flaws?

The research team consisted of Harsh Jaiswal, founder of Hacktron AI; AI security researcher Mohan Pedhapati; and computer scientist Rahul Maini.

The researchers said their work was conducted as security research rather than an attempt to misuse the access they discovered.

Their investigation focused on systems connected to OpenAI accounts and ultimately revealed weaknesses involving a third-party platform used by the company.

The incident occurred in July, with the researchers later sharing details of their findings.

How Claude Assisted the Security Research

The team used Anthropic’s Claude models during different stages of the investigation.

Their initial focus was the OpenAI community forum, where ChatGPT and Codex users can communicate and authenticate using OpenAI accounts.

The researchers used Claude Opus 4.8 to examine the code of Discourse, the third-party platform powering the forum. Initial attempts to demonstrate the suspected weakness were unsuccessful.

The researchers then turned to Claude Opus 5, which had been released later that day. According to the team, the newer model was able to help them successfully demonstrate the vulnerability.

By the next day, they said they could access ChatGPT and Codex accounts belonging to some users of the community forum.

Access Extended Beyond OpenAI Accounts

The researchers said the compromised account access could potentially provide pathways into other applications connected to the same identities.

These included services such as email and Slack. Some affected accounts reportedly belonged to OpenAI employees and were linked to internal company services.

According to the researchers, access could potentially expose users’ ChatGPT conversations, including private or sensitive information.

The team subsequently discovered another weakness involving OpenAI’s single sign-on system. They said the issue enabled them to obtain authentication information associated with ChatGPT and Codex accounts belonging to OpenAI employees.

The researchers also reported gaining access to an internal OpenAI code repository.

AI Significantly Reduced the Research Time

One of the most notable aspects of the incident was the amount of time required to complete the research.

Pedhapati estimated that conducting comparable work without Claude could have taken approximately two to three months.

With AI assistance, however, the team said it completed the research in less than three days.

He suggested that newer AI models could potentially reduce the time required even further, with similar research potentially taking less than a day under some circumstances.

The researchers described this capability as a force multiplier, allowing experienced security professionals to perform complex analysis and vulnerability research considerably faster.

Third-Party Software Created an Indirect Attack Path

The research also demonstrates why organizations cannot focus exclusively on vulnerabilities in their own software.

The researchers reportedly gained access through a weakness in Discourse, a third-party service used by OpenAI for its community forum.

This highlights the security risks created by interconnected software ecosystems. An organization may have strong security controls around its primary infrastructure while remaining exposed through an external service, authentication provider or connected application.

Third-party dependencies can therefore become an important part of an organization’s overall attack surface.

AI Could Accelerate Both Defensive and Malicious Research

The incident raises broader questions about the role of advanced AI in cybersecurity.

AI models can help security researchers analyze code, identify weaknesses and automate parts of vulnerability research. However, similar capabilities could also potentially be used by malicious actors to speed up attacks.

The researchers warned that increasingly capable AI systems could make experienced attackers more effective while lowering some of the technical barriers faced by less-skilled threat actors.

This creates a difficult security environment in which defenders may gain new capabilities at the same time as attackers gain access to similar technological advantages.

Connected Accounts Are Becoming a Critical Security Risk

The case demonstrates that compromising a single account or third-party application can have consequences beyond the original system.

Employee identities, single sign-on systems, cloud applications, messaging platforms, source-code repositories and other connected services can create chains of access.

For technology companies handling sensitive user information and proprietary systems, securing these connections is therefore as important as protecting the core application itself.

AI Is Changing the Economics of Security Research

The speed reported by the Hacktron AI researchers illustrates how AI could change the economics of vulnerability discovery.

Tasks that traditionally required substantial manual analysis and experimentation may increasingly be completed in a fraction of the time with capable AI assistance.

For defenders, this could mean faster vulnerability discovery and remediation. For attackers, the same acceleration could potentially shorten the window between identifying a weakness and attempting to exploit it.

As AI models become more capable, organizations may need to reassess how quickly they can detect, investigate and respond to newly discovered vulnerabilities.

Patna Cyber Police have busted an alleged organised cybercrime racket involved in digital arrest scams, financial fraud and online cheating, arresting three accused. Police said the suspects allegedly targeted victims across several states and were linked to a wider cybercrime network.

During the operation, investigators recovered six mobile phones, two laptops, three cheque books, two passbooks and five ATM cards, along with other electronic devices and banking documents.

Three Accused Arrested

The arrested accused have been identified as Gaurav Kumar, a resident of New Colony in Khagaul-Danapur; Adarsh Kumar of Transport Nagar; and Aman Raj of Kumhrar. According to police, the three were allegedly connected to an organised cybercrime network operating across different parts of the country. Preliminary investigation also revealed multiple complaints related to cyber fraud against them.

Cybercrime Complaints Helped Trace Suspects

Police said several complaints concerning the accused were found on the National Cybercrime Reporting Portal. Analysis of these complaints, along with digital evidence collected during the investigation, helped investigators trace the suspects and establish their alleged links with the racket.

Technical surveillance and examination of digital evidence played a key role in the investigation. Investigators analysed suspected mobile numbers, electronic devices and other digital activities before identifying and tracing the accused. The seized mobile phones and laptops have now been sent for detailed examination to identify additional suspects, victims and possible financial transactions.

Phones, Laptops And Banking Documents Seized

Police suspect that bank accounts and documents recovered from the accused may have been used to receive and transfer money obtained through cyber fraud. The three cheque books, two passbooks and five ATM cards are being examined to establish the flow of funds and identify accounts allegedly linked to the racket. Investigators are also examining the electronic devices for communication records, transaction details and other digital evidence.

How Digital Arrest Scams Work

The investigation is also focused on determining how the accused allegedly carried out digital arrest scams. In such frauds, cybercriminals typically impersonate police officers, central investigation agencies, bank officials or other government authorities and threaten victims with arrest or legal action. The victims are then kept under constant pressure through phone or video calls and are allegedly persuaded to transfer money on the pretext of verification, investigation or avoiding arrest.

FIR Registered, Investigation Expanded

An FIR has been registered at Patna Cyber Police Station, and the investigation has been expanded to determine the complete structure of the alleged network. The three accused were produced before a court and remanded to judicial custody. Police are now examining the seized electronic devices and banking records to establish the roles played by the accused and identify other members of the network.

Investigators are also trying to determine the total amount allegedly siphoned off by the racket and the number of victims who may have been targeted. Since complaints have surfaced from different parts of the country, authorities are examining whether the suspects were operating independently or as part of a larger network involving multiple handlers, bank accounts and money-transfer channels.

Expert Warns Against Digital Arrest Scams

Cybercrime expert and former IPS officer Prof. Triveni Singh said digital arrest scams combine technology with psychological manipulation. Victims are often kept under intense pressure and made to believe that immediate action is required to avoid arrest or other legal consequences. People should independently verify any such claim through the official contact details of the concerned department and should never transfer money merely because someone threatens them with arrest over a phone or video call.

Patna Cyber Police said the investigation remains underway. Further action will follow after the roles of other suspected members are established. Investigators are also analysing the digital and banking evidence recovered during the operation to determine the actual scale of the alleged fraud, trace the money trail and identify additional victims and associates connected to the network.