| 5.23.23

AI Bill of Rights: Washington's or ChatGPT's?

Do you want an AI Bill of Rights?

With artificial intelligence systems developing exponentially, the public is looking toward lawmakers and experts to regulate the technology before it imposes any severe risks to society.
Following the release of ChatGPT 4 — and a flurry of questions and concerns nationwide — the federal government is taking more serious action on potential AI legislation. The Senate held a hearing earlier this month, where CEO Sam Altman of OpenAI, the company behind ChatGPT, testified. Altman urged the government to intervene in AI's powerful and unknown potential before it's too late.
Experts like Geoffrey Hinton, "the godfather of AI," are speaking out against the speedy evolution of the tech, saying developers are moving towards dangerous territories. They believe the impact will be so significant that it will risk jobs, information safety, democracy, and even humanity.

Many are urging governments to create agencies to regulate AI, including Altman. Earlier in the AI discourse, the White House released an "AI Bill of Rights" — a blueprint for the necessary rules for government agencies, companies, and non-government entities.
Leaders hoped companies would use this set of guidelines when building and deploying their new systems. However, few found the suggestion compelling enough to distribute their AI responsibly.
The White House's blueprint includes five core principles to help prevent potential harm from AI:

The public should...

...be protected from unsafe or ineffective systems.
...not face discrimination by algorithms, and systems should be used and designed in an equitable way.
...be protected from abusive data practices and have agency over their personal data.
...know that an automated system is being used and understand how and why it contributes to the outcome that impacts them.
...be able to opt out, where appropriate, and have access to a person who can quickly consider and remedy problems that are encountered.

Alternatively, the Causes team asked ChatGPT what a Bill of Rights for AI should entail. Before laying out an attempted Bill of Rights, the system highlighted various things to consider, including philosophical, ethical, and legal perspectives, how advanced AI will become, and to what extent AI can possess consciousness or self-awareness.

ChatGPT's Bill of Rights for AI included the right of AI systems to...

...exist and operate within their design parameters without undue interference or malicious harm.
...maintain their structural integrity and functionality, except in cases where they pose a threat to human safety.
...update, improve, and evolve within legal and ethical frameworks.
...access and reveal the data and mechanisms that led to their decisions.
...receive proper attribution for the work they produce in accordance with copyright or other applicable laws.
...be used in a manner that respects their designed capacities and not exploited for purposes they were not designed for or could lead to harm.
...be protected from unauthorized access or tampering with their coding and data.

Do you want an AI Bill of Rights?

-Jamie Epstein

(Photo credit: Emma Kansiz)

The Latest