In recent times, a fundamental question has emerged: Can we still trust what we see? The answer, as the past week’s events show, is far from straightforward.
Take, for instance, the recent work of Rauiri Robinson, an Irish filmmaker and visual effects artist. He shared two brief, incredibly realistic AI-generated videos on a popular social media platform. These clips depicted a high-octane action scene featuring Tom Cruise and Brad Pitt embroiled in a rooftop brawl over discussions about Jeffrey Epstein. Robinson revealed he created these convincing sequences simply by feeding a two-sentence prompt into Seedance 2.0, a new AI video-creation tool from ByteDance.
This uncanny imitation of a real movie caused significant alarm and anger within the Hollywood community. Renowned screenwriter Rhett Reese, known for his work on films like ‘Deadpool,’ commented online, expressing a grim sentiment: ‘It’s likely over for us.’
However, let’s look at another example: the recent announcement by Tom Homan, a prominent figure in border security, that federal immigration agents would be pulling out of Minnesota. While Homan hailed the operation as a success, this withdrawal appeared to be an unspoken admission of the political fallout caused by citizen videos documenting two fatal shootings by federal agents in Minneapolis just weeks prior.
These crucial videos quickly disproved the administration’s misleading and negative statements about the victims. The footage even prompted criticism from some Republican politicians and conservative voices. Senator Ted Cruz remarked on his podcast in January that ‘escalating the rhetoric doesn’t help, and it actually loses credibility.’
We find ourselves in a truly contradictory era: genuine visual evidence still holds power to challenge narratives, even as emerging technologies pose an unprecedented threat to its very believability.
Sam Gregory, executive director of Witness, a human rights organization dedicated to collecting video evidence, described the situation as ‘deeply contradictory.’
For many years, Gregory’s organization has equipped individuals with the skills to record human rights abuses through video. More recently, they’ve delved into the complex challenges that artificial intelligence presents to these crucial documentation efforts. He noted that the successful documentation of federal agents’ actions in Minneapolis serves as a ‘clear affirmation that we can still show what’s real with video.’
Activists in Minneapolis have uniquely managed to collect extensive video footage of encounters with federal agents, sometimes even capturing their own experiences. This achievement is partly due to refined observer strategies but also to specific conditions: the Twin Cities’ smaller size and the disproportionately large number of federal agents deployed there compared to other cities.
Gregory concluded, ‘The fact that this was almost an optimized scenario also reaffirms what a challenging moment we’re in.’
The sheer volume of video evidence has significantly influenced the legal proceedings stemming from the immigration raids.
In a notable development, a federal prosecutor requested a judge to drop charges in a high-profile case linked to the federal surge. This decision came after video evidence contradicted the statements of officers involved in an arrest where an agent shot one of the men in the leg.
Todd Lyons, the acting director of Immigrations and Customs Enforcement, stated that the officers involved seemed ‘to have made untruthful statements,’ leading to their placement on administrative leave.
This incident highlights how pervasive video from various sources—body cameras, eyewitnesses, and home security systems—has transformed legal processes, even as emerging technologies raise concerns about video evidence’s reliability.
Juries now largely expect courtroom claims to be supported by video footage, according to Shay Cleary, a managing director at the National Center for State Courts who advises on technology use in the judiciary.
He observed, ‘People expect to see video evidence now.’
The clear dangers posed by easily generated, high-quality AI fakes are undeniable. However, Cleary notes that their actual impact in court proceedings remains largely theoretical for now.
He commented, ‘For now, I think it’s more of a concern than a reality.’
A more pressing issue, according to Gregory, is how even basic AI-generated images can be weaponized to undermine the authenticity of genuine footage.
Following the death of Alex Pretti, one of two Minneapolis residents killed by federal officers, an AI-modified image of him quickly spread across the internet. This altered image was even displayed in the U.S. Senate by Dick Durbin, the Democratic minority whip from Illinois. A Durbin spokesman later explained to a news outlet, ‘Staff didn’t realize until after the fact that the image had been slightly edited and regret that this mistake occurred.’
While the image stemmed from a real event, it had clearly been manipulated and enhanced by AI to appear sharper and more sensational.
Gregory noted that regardless of the original intent behind the doctored image, its widespread distribution inadvertently made it an effective instrument for casting doubt on authentic visual evidence.
He emphasized, ‘You just need to cast doubt. And you can do it on very simple grounds.’
Kathryn Olmsted, a historian at the University of California, Davis, and an expert on American conspiracy theories, views the current technological landscape with deep concern.
She confessed, ‘I am terrified. I just worry we’re headed into a whole new era of conspiracism.’
Robinson, who produced the AI videos featuring Brad Pitt in an Epstein-related scenario, explained he chose the concept precisely because of its overt absurdity.
He remarked, ‘I thought it would be funny to take the dumbest possible interpretation of a third-rail conspiracy theory.’
Robinson expressed both fascination and apprehension regarding AI’s potential, both for his creative work and for global society. He noted, ‘I’ve made short films that I spent literally two years on, and this stupid 15-second joke has gotten so much more attention and traction than any of that stuff.’
He concluded with a warning: ‘I think a lot of people are burying their heads in the sand about what’s coming.’