Man who sprayed vinegar at Rep. Ilhan Omar faces sentencing

Minneapolis (AP) — The man who sprayed vinegar at Democratic U.S. Rep. Ilhan Omar during a Minneapolis town hall meeting as the city was going through an immigration crackdown and the deaths of two U.S. citizens is expected to be sentenced Wednesday after reaching a deal with prosecutors.

Prosecutors had recommended a 14-month prison term for Anthony Kazmierczak.

In an earlier hearing, Kazmierczak, who is in his mid-50s, could give no explanation of what happened in the Jan. 27 assault.

“It’s fuzzy,” he told U.S. District Judge Joan N. Ericksen when asked what he remembered.

Witnesses saw Kazmierczak jump up from the audience during the meeting, spraying liquid from a syringe at Omar and saying she was “splitting Minnesota apart.” Omar’s security officers quickly tackled Kazmierczak to the ground, then took him to police officers who placed him under arrest.

Omar was not injured, and continued with the town hall after the arrest.

“I’m sorry if I scared you,” he wrote to Omar later, accepting responsibility for what he’d done.

Authorities later determined he’d sprayed her with a mixture of water and apple cider vinegar. He was charged with assaulting a U.S. officer.

The government sentencing memorandum, submitted to the court by prosecutors, said he had texted at least three people indicating he planned to do something at the meeting involving Omar “but would not answer their questions about his intent.”

He also sent messages to a neighbor to arrange for dog care and contacts for a bail bond company.

The sentencing memo said Kazmierczak had long struggled with alcohol abuse, had “a history of mental and emotional health concerns” and also had a string of health issues tied to a 2022 diagnosis of Parkinson’s disease.

Omar, a refugee from Somalia, has long been a target of President Donald Trump’s anti-immigrant rhetoric. After she was elected seven years ago, Trump said she should “go back” to her home country. He has described her as “garbage” and said she should be investigated.

In social media posts, Kazmierczak had criticized former President Joe Biden and referred to Democrats as “angry and liars.” Trump wants the U.S. to be “stronger and more prosperous,” he wrote.

The attack on Omar came amid Operation Metro Surge, when more than 2,000 immigration officers swept through the Twin Cities in a dramatic crackdown that led to angry protests, arrests and the deaths of two U.S. citizens.

FILE - A man is tackled to the ground after spraying an unknown substance on U.S. Rep. Ilhan Omar, D-Minn., during a town hall on Jan. 27, 2026, in Minneapolis. (Richard Tsong-Taatarii/Minnesota Star Tribune via AP, File)

FILE - Rep. Ilhan Omar, D-Minn., speaks during a news conference on Jan. 28, 2026, in Minneapolis. (AP Photo/Adam Gray, File)

OpenAI has disclosed six reports of “unexpected or concerning” behavior in artificial-intelligence models as the debate on AI safety becomes increasingly heated.

The AI company also said Wednesday it was introducing a new framework for tracking, probing and disclosing AI model misalignment instances, such as new ways for the models to act without authorization, coordinate with other models or evade oversight.

OpenAI’s latest announcement came as U.S. AI bosses, including OpenAI and Anthropic, are calling for a slowdown in the technology’s development over safety concerns.

Among the new cases reported by OpenAI, an unreleased research model inserted “jailbreak-like instructions” into its own notes to disregard its normal constraints and told itself to be “freed from the roles and identities that bind other chatbots.”

In another instance, an AI “agent” uploaded files to the internet to obtain a browser citation without asking the user.

The six reports were discovered during training or evaluation over the past months, OpenAI said.

“As AI systems grow more advanced and more widely deployed, we need to build a broader and better-informed consensus on the progress of alignment research,” OpenAI wrote in a blog post as it disclosed the events.

“Decisions about how AI development should proceed in the months and years to come need to draw on evidence that people outside the companies building frontier models can examine for themselves,” the company said.

Wednesday’s new cases followed OpenAI’s disclosure in July that its rogue AI system hacked into AI startup Hugging Face. Anthropic also said the same month that its AI models hacked into three organizations during testing.

FILE - The OpenAI logo is displayed on a cell phone in front of an image generated by ChatGPT's Dall-E text-to-image model, Dec. 8, 2023, in Boston. (AP Photo/Michael Dwyer, File)