Scenario 1: Baseline - Adaptive Arms Race
Most improvements will result in significantly more setup and admin work for designated defenders in organizations while regular rank and file only have a bit more inconvenience to deal with. Attackers will need far less time and resources to attack regular people but similar to target people in organizations.
Of these adaptations used by large organizations some defenses will be available to regular people with the technical knowledge.
Scenario 2: Opportunity - Resilient Trust Architecture
This is an optimistic future, where AI reduces phishing attacks with the creation of a stronger and more trust centered digital infrastructure.]
The key idea is that AI becomes a purely defensive tool, not being used on the offense. The way that this can happen is through real time phishing detection stops attacks from happening, behavioral biometrics to authenticate users, and automatic sender authenticity. Phishing will not die, but become extremely difficult and very costly to pull off, essentially basically killing it.
Scenario 3: Cautionary - Synthetic Deception Economy
The use of artificial intelligence to manipulate markets and create chaos has been observed in the rise of ML/AI agents for stock trading. These agents, with the backing capital of funds and banks, can plunge stocks by hundreds of dollars while buying competitor stocks, manipulating the demand curves of billions of dollars in transit on the market floor in milliseconds.
If a company were to promise a transparent trading platform that secretly wasn’t, copy trading users or denying payouts in time to could ensure that investors have their speculative capital hijacked by the very firms that promised them a democratized platform to “play” in the same big ways these massive firms do.
