Imagine Michael Jackson doing banana-based comedy or Bruce Lee running a DJ set�scenes that never happened in reality but are now flooding social media thanks to OpenAI’s Sora 2 video generator? While these AI-generated videos of deceased celebrities spark amusement and controversy, they reveal a deeper, more troubling trend: the rapid expansion of AI capabilities without adequate safeguards? As companies race to integrate AI into every facet of business, from creative tools to corporate workflows, the risks are multiplying faster than solutions can emerge?
The Digital Afterlife: When AI Resurrects the Dead
OpenAI’s Sora 2, launched last week, includes measures to block depictions of living public figures by default? However, as detailed in primary source reporting, this protection doesn’t extend to historical figures? The result? A flood of AI-generated videos featuring deceased celebrities like Tupac Shakur, Stephen Hawking, and Martin Luther King Jr? in scenarios ranging from absurd to disrespectful? Zelda Williams, daughter of the late Robin Williams, publicly pleaded for people to stop sending her AI videos of her father, stating, “It’s dumb, it’s a waste of time and energy, and believe me, it’s NOT what he’d want?”
This raises critical questions about consent and control in the AI era? While living individuals can use Sora 2’s “cameos” feature to control their likeness, deceased public figures have no such protection? OpenAI acknowledges this gap, with a spokesperson telling PCMag, “We don’t have a comment to add, but we do allow the generation of historical figures?” The legal landscape remains murky�while states like California and New York have right-of-publicity laws, court rulings often protect “transformative” uses under the First Amendment?
Beyond Entertainment: The Corporate Security Nightmare
While AI-generated celebrity videos capture headlines, the real business impact lies in the proliferation of AI agents within corporate environments? According to research from the OpenID Foundation (OIDF), unchecked AI agents could create catastrophic security vulnerabilities? The problem isn’t theoretical�imagine an employee granting an AI agent access to their email inbox to automate customer responses? Today, this might involve a handful of early adopters, but within five years, every employee could have multiple AI agents operating with minimal oversight?
“Whereas employees once outnumbered the agents, suddenly the agents�all with a wide variety of human-like access to corporate resources�will outnumber the employees,” warns the OIDF research? The Model Context Protocol (MCP), which enables AI agents to dynamically access diverse resources, becomes a “double-edged sword”�enhancing capabilities while making agents less predictable and harder to control? As paper author Tobin South told ZDNET, “MCP’s IAM controls are a start, but they’re not nearly robust enough for the expanding surface area?”
The Music Industry’s AI Invasion
The creative industries are already feeling the impact? Spotify removed 75 million “spammy” AI-generated tracks in the past year, while French streaming service Deezer reports that over 28% of daily uploads are now AI-generated, up from just 10% in January? This isn’t just about quantity�AI is now producing content difficult to distinguish from human artists? The indie rock band The Velvet Sundown amassed millions of streams before being revealed as entirely AI-generated?
For the $105 billion music industry, this represents an existential threat? While copyright laws may protect against outright fraud, the larger concern is marginalization of human artists as AI floods platforms with humanlike music? The industry response is evolving�labels may create their own AI artists while emphasizing live events, which already account for about a third of global music revenue? Universal Music and Warner Music are reportedly considering licensing deals with AI companies, but the fundamental question remains: Can they avoid selling their artists’ work for a song?
Building Guardrails for the AI Revolution
The OIDF research identifies urgent needs for AI governance, including giving AI agents “first-class identity considerations” similar to human users? This means extending identity and access management (IAM) controls to AI systems, with specialized guardrails for their unique characteristics? The research suggests using protocols like SCIM (System for Cross-domain Identity Management) to automate lifecycle management of AI agents, ensuring they’re governed by the same policy-driven workflows used for human employees?
These guardrails can “help prevent unintended behaviors, reduce risks, and maintain trust by guiding AI agents to act responsibly and in alignment with human values,” according to the OIDF paper? The challenge is implementing these controls before the problem scales beyond management? As the research notes, “Hope is not a strategy” when dealing with exponentially growing AI deployments?
The Broader Business Implications
These developments occur against the backdrop of massive AI infrastructure investments? OpenAI has signed approximately $1 trillion in computing power deals this year alone, securing over 20 gigawatts of computing capacity�equivalent to 20 nuclear reactors? Each gigawatt costs about $50 billion to deploy, raising questions about how a company expected to lose $10 billion this year will fund these commitments?
The financing arrangements are increasingly complex? AMD granted OpenAI warrants for up to 160 million shares, with vesting tied to stock price milestones? As UBS analyst Timothy Arcuri notes, if OpenAI holds the stock until the end of the deal, its stake could be worth about $100 billion? But more likely, OpenAI will sell shares along the way to pay its AMD bill�essentially using AMD’s own stock to finance its purchases?
This circular financing highlights the speculative nature of current AI investments? As analyst Gil Luria of DA Davidson observes, “OpenAI is in no position to make any of these commitments? Part of Silicon Valley’s ‘fake it until you make it’ ethos is to get people to have skin in the game? Now a lot of big companies have a lot of skin in the game on OpenAI?”
Navigating the New AI Landscape
For businesses and professionals, the message is clear: AI’s potential must be balanced against its risks? The same technology that can resurrect celebrities for entertainment can also create security vulnerabilities that threaten corporate integrity? The tools that enable creative expression can also flood markets with synthetic content that devalues human creativity?
The solution lies in proactive governance rather than reactive measures? Companies need to extend their security frameworks to include AI-specific controls, treating AI agents as first-class citizens in their identity management systems? Industries must establish clear guidelines for AI-generated content that protect intellectual property while allowing for innovation? And regulators need to update legal frameworks to address the unique challenges posed by AI, from digital likeness rights to agent accountability?
As OpenAI CEO Sam Altman acknowledged regarding Sora 2’s handling of copyrighted works, “We will make some good decisions and some missteps, but we will take feedback and try to fix the missteps very quickly?” The question for businesses is whether they can afford to wait for those missteps to be corrected, or if they need to build their own guardrails now?