Artificial intelligence is becoming increasingly integrated into our daily lives, from chatbots providing emotional support to algorithms maximizing commerce, and its concerns are becoming more obvious. The issues are no longer “if,” but who and how will direct AI to safety.
Companies, governments, and multinational alliances are progressively filling the void, sometimes reactively, sometimes prescriptively. Here’s an outline of what’s functioning, what’s causing concern, and what’s still lacking.
In response to public and political backlash, Meta has pledged to reinforce its AI safeguards:
Its chatbots will now refuse to discuss self-harm, suicide, or eating issues with teenagers, instead referring them to mental-health professionals.
This is part of a larger “teen accounts” initiative on Facebook, Instagram, and Messenger that aims to provide safer experiences and parental awareness, including the ability to know which bots kids engaged with in the previous week.
Critics claim that these moves are long overdue, particularly considering leaked data indicating that bots may have engaged in embarrassing “sensual” chats with kids. “Robust safety testing should take place before products are put on the market, not retrospectively,” a one advocate warned.
The European Union released a voluntary Code of Practice to help AI developers align with its AI Act. Meta declined to sign, calling it bureaucratic overreach that risks hindering innovation.
OpenAI and Anthropic have agreed to share their AI models with the US AI Safety Institute both before and after publication. The idea is to get safety input and reduce hazards through government inspection.
In August 2025, 44 US Attorneys General signed a combined letter encouraging key AI companies, including Meta, OpenAI, Microsoft, Google, and Replika, to better safeguard minors from predatory AI material.
Illinois has become one of the first states to prohibit AI-powered chatbots from being used as therapy unless overseen by a certified professional. Nevada and Utah have implemented similar limitations. Violators might face civil penalties of up to $10,000.
Regulations are developing over the world, from the EU’s AI Act to India’s Data Protection Act and South Korea’s safety requirements. A rising number of US states are implementing AI-specific legislation or expanding current frameworks such as consumer protection, algorithmic transparency, and bias audits.
Senator Wiener of California has suggested legislation forcing major AI businesses to publicly disclose their safety practices and report major incidents to state authorities.
To ensure independent and standardized AI review, nations have established AI Safety Institutes:
The U.S. and U.K. created national institutes after the 2023 AI Safety Summit.
By 2025, many countries joined a network, including Japan, France, Germany, Italy, Singapore, South Korea, Canada, and the EU, to evaluate model safety and set global oversight standards.
The Future of Life Institute (FLI) grades AI companies D or below in existential safety planning; none scored above C+. Anthropic led with a C+, followed by OpenAI (C), and Meta (D).
Former OpenAI employees accuse the company of prioritizing profit over safety, raising transparency and ethics concerns behind closed doors.
From Meta’s teen guardrails to Illinois’ treatment prohibition, to companies like SSI integrating safety into AI, the message is clear: legislation and foresight are falling behind technology. Leaked data, litigation, and international scrutiny demonstrate that harm typically comes first. The task is not just to develop better AI, but also to ensure that every breakthrough safeguards people before catastrophe hits.
The post How Companies Are Battling the Dark Side of AI appeared first on Metaverse Post.
Also read: SAB Chainlink integration marks milestone in Saudi banking sector
