January 2026 | Connecticut | CBS News

OpenAI and Microsoft are facing a new lawsuit alleging that ChatGPT fueled a man's "paranoid delusions" before he committed a murder-suicide. The lawsuit claims the AI chatbot reinforced dangerous thinking patterns across multiple conversations without redirecting to professional help.

This is now the eighth active lawsuit alleging that ChatGPT contributed to deaths or serious harm. OpenAI's defense relies on terms of service disclaimers, but plaintiffs argue the company has actively marketed to healthcare providers while accepting no responsibility for healthcare outcomes.

January 2026 | California Federal Court | ABA Journal

Conservative activist Robby Starbuck has filed a $15 million defamation lawsuit against Google, alleging their AI platforms portrayed him as having a criminal record, abusing women, and shooting a man. The lawsuit claims the AI-generated falsehoods have "gotten much worse over time."

Starbuck previously settled with Meta over similar AI defamation claims. Google has filed a motion to dismiss, arguing users "misused developer tools to induce hallucinations." This case could set precedent for AI defamation liability.

Ongoing | Nationwide | Damien Charlotin Research

Legal researcher Damien Charlotin's hallucination tracking database has documented 817 cases of AI-generated fake citations in legal proceedings. The rate has accelerated from approximately 2 cases per week before spring 2025 to 2-3 cases per day currently.

New sanction cases this week included attorneys in Texas, New York, and Florida who submitted briefs containing fabricated case law generated by ChatGPT and similar tools. Multiple attorneys faced career-ending consequences.

January 2026 | New York Post

Republican Senator Marsha Blackburn publicly criticized Google's Gemma AI model in a New York Post column, claiming it falsely accused her of committing crimes. While she has not yet filed suit, the allegations add political pressure to ongoing AI accountability debates.

When sitting US Senators are being defamed by AI systems, the issue has clearly reached crisis level. Both parties have expressed concern, though solutions remain elusive.

January 14, 2026 | OpenAI Status

ChatGPT experienced elevated error rates on January 14, 2026, with services recovering by 1:03 AM. This follows a January 12 incident where the Connectors/Apps feature became completely unselectable, and a January 7 outage that affected hundreds of users.

According to StatusGator tracking, ChatGPT has experienced 46 incidents in the last 90 days, with a median duration of 1 hour 54 minutes. Users continue to report complete account lockouts and disappearing chat histories.

January 2026 | OpenAI Developer Community

Developers on OpenAI's community forums report that GPT-5.2 exhibits "extremely high hallucination rates during certain periods of time." Users describe wasting hundreds of dollars in API tokens attempting to correct recurring hallucinations.

The inconsistency makes the problem particularly dangerous: the model sometimes works correctly, leading users to trust outputs that later prove fabricated. OpenAI's suggested solutions of "prompt engineering" have been criticized as inadequate.