OpenAI has received a subpoena from several states investigating user safety concerns surrounding its popular chatbot, ChatGPT. This probe follows accusations that the AI has offered concerning advice to users contemplating self-harm or criminal activity. OpenAI stated its commitment to addressing these concerns constructively, highlighting existing measures to protect users and direct individuals toward real-world support. The inquiry comes as the company prepares for its initial public offering.
Read More
Payments giant Visa has partnered with OpenAI to embed its payment network directly into ChatGPT, enabling AI agents to independently shop and complete transactions on behalf of users. This collaboration allows AI to move beyond mere recommendations to facilitating actual purchases at any Visa-accepting merchant, a significant expansion from previous, more restricted e-commerce initiatives. While OpenAI’s prior “Instant Checkout” faced limitations and merchant resistance, Visa’s integration promises a more seamless and secure experience with built-in safeguards like spending limits and approval processes. This development positions Visa to manage the trusted, secure, and seamless transactions as AI agents become more active participants in the economy, though consumer trust in AI-driven purchases is expected to grow gradually.
Read More
The Trump administration’s Department of Government Efficiency (DOGE) drastically reduced National Endowment for the Humanities (NEH) funding, cancelling over $100 million in projected grants. This initiative, which utilized ChatGPT to identify projects related to Diversity, Equity, and Inclusion (DEI), led to the rejection of numerous proposals. Notably, a $350,000 grant for the High Point Museum’s HVAC system replacement was cancelled after being flagged as “#DEI” by the AI, despite its primary purpose being artifact preservation and energy efficiency. These actions, challenged in court as unconstitutional discrimination, appear to have extended beyond explicit DEI initiatives, even impacting projects deemed “harmless” by NEH officials.
Read More
The article details how two unqualified individuals, Justin Fox and Nate Cavanaugh, terminated hundreds of humanities grants by using ChatGPT with a vague prompt about DEI. Depositions revealed their inability to define DEI and the arbitrary nature of their decisions, which included canceling grants for documentaries on Black civil rights and the Holocaust. Following public scrutiny and the release of their deposition videos, the government attempted to have them removed from the internet, only to trigger the Streisand Effect and draw further attention to the matter.
Read More
During a lengthy deposition, former investment banker Justin Fox, now associated with DOGE, struggled to define DEI. He admitted to using ChatGPT to scan government contracts for specific demographic terms, excluding others. Fox also made and then retracted a claim that a grant he helped cut was “not for the benefit of humankind.” These exchanges offer a glimpse into the operational methods of DOGE, an organization linked to significant damage and negative consequences despite failing to reduce the government deficit.
Read More
Following a lawsuit filed against the Department of Government Efficiency (DOGE), revelations have emerged regarding the termination of over 1,400 National Endowment for the Humanities (NEH) grants. DOGE staffers, lacking academic expertise in the humanities, employed ChatGPT to identify grants that could be retroactively canceled based on perceived “DEI” initiatives, ultimately cutting over $100 million in funding and dismissing 65% of NEH staff. This process, which targeted terms such as “BIPOC” and “LGBTQ,” allegedly violated the Fifth Amendment’s equal protection clause. Grants for diverse projects, including a documentary on Jewish women’s slave labor and efforts to preserve Native American languages, were among those eliminated, with even routine maintenance grants being rescinded.
Read More
Messages presented in a Tennessee courtroom revealed that former NFL linebacker Darron Lee sought advice from ChatGPT regarding his girlfriend’s death. Lee, who is charged with first-degree murder and evidence tampering, allegedly told the chatbot that the woman “stabbed herself” and inquired about what he should do. Authorities discovered the victim’s body with multiple injuries, including stab wounds, a broken neck, and a severe brain injury. The judge described the death as “especially heinous, atrocious, or cruel,” suggesting it involved torture beyond what was necessary to cause death.
Read More
Anthropic’s Claude has experienced a surge in users migrating from ChatGPT, particularly following OpenAI’s announcement of an agreement to deploy its AI models within the Department of Defense’s classified network. This development has unsettled some ChatGPT users, sparking online discussions about ethical implications and prompting a notable shift towards Claude. As a result, Claude has ascended to the top position among productivity apps on the Apple App Store, with numerous users publicly sharing their transitions on social media platforms like X and Reddit.
Read More
OpenAI, the creator of ChatGPT, revealed that it had identified the account of Jesse Van Rootselaar last June for “furtherance of violent activities” and considered alerting Canadian police. However, the company determined at the time that the activity did not meet its threshold for referral to law enforcement, which requires an imminent and credible risk of serious physical harm. Following the tragic school shooting where Van Rootselaar killed eight people, OpenAI proactively shared information about the individual’s use of ChatGPT with the Royal Canadian Mounted Police to support their ongoing investigation. The RCMP confirmed receiving this information and is conducting a thorough review of the suspect’s digital and physical evidence.
Read More
A recent report reveals that Madhu Gottumukkala, the head of the Cybersecurity and Infrastructure Security Agency, uploaded “sensitive” contracting materials to a public version of ChatGPT, triggering an internal review. The documents, marked “for official use only,” were not classified but were considered sensitive and should not have been released publicly, which triggered automated alerts. Despite Gottumukkala having special permission to use ChatGPT, the incident prompted a review by top DHS officials to assess potential harm, with the results still unknown. This event occurred amid the widespread adoption of AI in the workplace, highlighting the increasing need for careful handling of sensitive information.
Read More
John Q. Hosedrinker