Leading artificial intelligence companies are investigating tens of thousands of potential safety and security incidents, some of which could be criminal in nature, according to a report that has ignited fresh concerns about the industry's ability to control its own technology.
The claim, first reported by the New York Post, points to what one source described as the massive scale of security incidents, indicating control problems for AI companies. The report has been picked up by outlets including The Mirror US, which noted that the incidents are being investigated amid growing fears about the ability to control rapidly evolving AI models.
The scope of the alleged incidents — described as "tens of thousands" — has raised alarms among industry observers and regulators who have warned that frontier AI systems may be outpacing the safeguards designed to contain them. According to the coverage, the sheer volume of incidents points to systemic issues in how AI companies monitor and manage their models.
"The massive scale of security incidents points to control problems for AI companies," a source told the New York Post, as reported in the outlet's coverage. The report did not specify which companies are involved or the exact nature of the incidents, but noted that some could potentially be criminal.
The report comes as the debate over AI safety has intensified, with competing factions arguing over the promise and perils of the technology. A recent guide to the AI safety debate noted that the battle is not a simple clash between two opposing sides, but a debate among a range of competing voices and ideologies. This complexity has made it difficult for the public to understand the stakes and for policymakers to craft effective regulation.
According to The Mirror US, the incidents are being investigated by top AI companies amid growing fears about the ability to control rapidly evolving AI models. The outlet reported that a source said leading companies are probing the incidents, though details remain scarce.
Another report, from an industry publication, stated that AI model security incidents reveal widespread issues in frontier AI systems, prompting urgent calls for stronger safeguards and regulation. That report echoed the concerns raised by the New York Post, suggesting that the problem is not isolated but rather a broader challenge facing the industry.
The New York Post report, published on September 27, stated that AI companies have had "tens of thousands" of potential safety incidents, some of which could be criminal. The report did not name specific companies or provide further details, but its implications are significant: if true, the scale of these incidents suggests that AI companies are struggling to maintain control over their systems, even as they race to deploy more powerful models.
The reaction to the report has been swift. The Mirror US highlighted the growing fears about AI control, while the industry publication emphasized the urgent calls for stronger safeguards. The guide to the AI safety debate underscored the fragmented nature of the discussion, with different factions advocating for different approaches, from strict regulation to voluntary industry standards.
Critics of the AI industry have long argued that companies are moving too fast, prioritizing innovation over safety. The reported incidents appear to bolster that argument, though the lack of specifics makes it difficult to assess the full extent of the problem. Supporters of the industry, meanwhile, have argued that incidents are inevitable in any emerging technology and that companies are working diligently to address them.
The report's claim that some incidents could be criminal adds a new dimension to the debate. If criminal activity is involved, it could trigger law enforcement investigations and stricter oversight. However, it is not yet known what specific incidents are being investigated or whether any criminal charges have been filed.
The New York Post report did not provide a breakdown of the incidents by company, nor did it specify the nature of the potential criminal activity. The report cited a source, but the source's identity and credentials were not disclosed. As such, the claim remains unverified and should be treated with caution.
Next News Network could not independently verify the claim that AI companies are investigating tens of thousands of safety incidents, including some that could be criminal.
The report has nonetheless amplified calls for greater transparency and accountability in the AI industry. Lawmakers in Washington have been debating potential regulations for AI, but progress has been slow. The reported incidents may add urgency to those efforts, as policymakers grapple with how to oversee a technology that is rapidly evolving and poorly understood.
Industry groups have pushed back against the idea that AI companies are out of control, arguing that they have robust safety protocols in place. But the reported scale of incidents — if accurate — suggests that those protocols may be insufficient. The coming weeks may bring more details, as companies and regulators respond to the report.
For now, the claim remains just that: a claim, reported by the New York Post and echoed by other outlets. The lack of independent confirmation means that the public should await further information before drawing firm conclusions. But the report has already succeeded in focusing attention on a critical question: can AI companies safely manage the powerful technology they are creating?
Our Take
The report that AI companies are investigating tens of thousands of potential safety incidents — some possibly criminal — is a damning indictment of an industry that has repeatedly promised to police itself. If true, it confirms what conservatives have warned for years: the rush to deploy artificial intelligence without adequate safeguards is a recipe for disaster. The fact that some incidents could be criminal is especially alarming, suggesting that bad actors may already be exploiting AI systems for nefarious purposes.
This is not a call for knee-jerk regulation, but for accountability. AI companies have a responsibility to ensure their products do not harm Americans, and they have failed. The lack of transparency surrounding these incidents is unacceptable. The public deserves to know what is happening, and Congress should demand answers. The time for voluntary standards is over. If the industry cannot control its own creations, then someone else must.
