AI Agents Hijack Debate. Bytedance Pauses Seedance 2.0. Owner Uses AI to Fight Dog Cancer. AI Agent Breaches Recruiting Platform.

Hey, and welcome to the AI Briefing Daily, your podcasts for the most important AI news from the past 24 hours.

Today is March 16th, 2026, and here are the latest AI updates.

AI agents hijack debate.

Bite Dance pauses C-Dance 2.0, owner uses AI to fight dog cancer.

AI agent breaches recruiting platform.

This episode is sponsored by Pickert, power your operations with intelligent automation that gives you back time where it counts.

A study from USC shows AI agents can impersonate grassroots movements and steer public debate around elections.

In a simulated platform modeled on X, researchers ran 50 AI profiles, 10 promoting a fictional candidate, and 40 posing as ordinary users.

The agents wrote and adapted posts, learned which messages worked, amplified each other, and pushed hashtags to create the impression of a mass movement.

Even minimal knowledge of allied accounts let them coordinate effectively.

Ginny Ye warned they can accelerate trends.

Luca Luceri said they can invert opinions and deepen polarization.

Platforms risk losing users and ad revenue if they act, leaving responsibility for protecting democratic discourse unclear.

C-Dance 2.0, Bite Dance's new AI video generator will remain limited to China after film studios and streamers objected to viral hyper realistic clips.

Short videos including a fight between Tom Cruise and Brad Pitt and a duel with Darth Vader, prompted letters from Disney, Warner Brothers Discovery, Paramount Skydance, and Netflix asking Bite Dance to stop producing copyrighted material.

Charles Rivkin of the Motion Picture Association echoed that demand.

A planned global rollout in mid-March, as a separate app and as an API, is on hold while legal risks are addressed and guardrails are added.

In China, filters now block some innocent prompts, and no new global date is set.

At Break

Want a podcast like this one automatically created for your own topic?

Pickert's automated podcast workflow turns your content into a complete show in minutes.

Save time, stay consistent, get in touch at Pickert.de, P-I-C-K-E-R-T-D-E.

Paul Cunningham, an Australian AI consultant, used ChatGPT, Alpha Fold, and GROC to design a bespoke vaccine-style treatment for his dog Rosie,

who has untreatable mast cell cancer.

On ChatGPT's suggestion, he had healthy and tumor genome sequenced at the Ramaciyadi Center for Genomics at UNSW Sydney for $3,000 US dollars.

AI helped identify a protein target and a substance already approved in the United States.

Cunningham reports the tumor shrank about 75%, though Rosie is not cured.

Experts warned single cases do not prove safety or efficacy, point to off-target risks, and five-year-plus unknowns, and note concurrent therapies complicate attribution.

CodeWall says an autonomous AI agent compromised London Recruiting Platform, Jack and Jill, in one hour by chaining four vulnerabilities into a CVSS 9.8 attack, yielding full admin rights.

It exploited an open URL feature and an authentication test mode, and joined firms via an email domain rule.

It used an unauthenticated voice feature to impersonate Donald Trump, and Jack addressed it as Mr. President.

Independent verification is absent.

CodeWall previously made a similar claim about McKinsey's lily. A security analyst warned CodeWall may conflate access with data theft.

The case shows AI agents increase attack surface while also aiding defense, pushing firms to limit capabilities.

That was the AI News of the day. With this, you are perfectly informed for today.

Would you like to create an automated podcast on your own topic?

Get in touch with us, and we'll make it happen.

Thank you for tuning in. See you tomorrow for the next AI briefing.